检索结果-内蒙古大学图书馆

Youth Academic Annual Conference of Chinese Association of Automation (YAC)

作者： Panpan Zhang Haibo Luo Zheng Xu Miao He Shenyang Institute of Automation Shenyang China Institutes for Robotics and Intelligent Manufacturing Shenyang China University of Chinese Academy of Sciences Beijing China Key Laboratory of Opto-Electronic Information Processing Shenyang China The Key Lab of Image Understanding and Computer Vision Shenyang China

ISBN: (纸本)9781665465373

At present, deep learning technology is widely used in ship target detection in synthetic aperture radar (SAR) images. However, high-resolution remote sensing SAR images cover a larger area and have larger image sizes. To be able to use the deep learning model for training and testing, the image needs to be cropped to the appropriate size. In a high-resolution SAR ship image, the ship target usually takes only a small part of the whole image. As a result, only part of the cropped image contains the target, and most of the rest are background regions. This will cause a lot of computational redundancy in the model inference stage. To solve this problem, a fast candidate region extraction algorithm is proposed in this paper for ship target extraction. The algorithm consists of three parts: firstly, extract the saliency map of the SAR ship image, secondly, perform median filtering on the segmented image, and thirdly, extract candidate regions. The superiority of the algorithm was demonstrated by experiments on the AIR-SARShip-1.0 dataset.

关键词： Deep learning Training Redundancy Radar polarimetry Inference algorithms Indexes Marine vehicles

来源：评论

学校读者我要写书评

暂无评论

Head-free, Human Gaze-driven Assistive Robotic System for Reaching and Grasping 40

Head-free, Human Gaze-driven Assistive Robotic System for Re...

引用

40th Chinese control Conference (CCC)

作者： Yang, Bo Huang, Jian Sun, Menglin Huo, Jun Li, Xiaolong Xiong, Caihua Huazhong Univ Sci & Technol Sch Artificial Intelligence & Automat Key Lab Image Proc & Intelligent Control Wuhan 430074 Peoples R China Huazhong Univ Sci & Technol Sch Mech Sci & Engn Wuhan 430074 Peoples R China Huazhong Univ Sci & Technol State Key Lab Digital Mfg Equipment & Technol Wuhan 430074 Peoples R China

ISBN: (纸本)9789881563804

Patients with limb dysfunction have limited mobility, which prevents them from performing daily activities. We have developed an assistive robot system with an intuitive head free gaze interface. The system consists of multiple modules, including 3D gaze estimation, head free coordinate transformation, intention recognition, and robot trajectory planning. The robotic assistive system obtains clues from the user's gaze to decode their intentions and implement actions. This allows the user only needs to look at the objects to make the robot system reach, grasp, and bring them to the user. The 3D gaze estimation is evaluated with 5 subjects, showing an overall accuracy of 5.53 +/- 1.2 cm. The integrated system's experimental results show that the success rate is 96% in the implementation of automatic trajectory planning, and the success rate is 92% in the implementation of fixation-based trajectory planning. Finally, the results and work required to improve the system are discussed.

关键词： 3D gaze estimation Intention estimation Gaze-based control Assistive robot

来源：评论

学校读者我要写书评

暂无评论

Named Entity Recognition Based on Anchor Span for Manufacturing Text Knowledge Extraction

SSRN

引用

SSRN 2024年

作者： Li, Yahui Sun, Qi Zhou, Chunjie Liu, Lu Tian, Yu-Chu School of Artificial Intelligence and Automation Huazhong University of Science and Technology Hubei Wuhan430074 China Key Laboratory of Image Processing and Intelligent Control Ministry of Education Huazhong University of Science and Technology Hubei Wuhan430074 China School of Cyber Science and Engineering Huazhong University of Science and Technology Hubei Wuhan430074 China School of Computer Science Queensland University of Technology BrisbaneQLD4001 Australia

intelligent industrial manufacturing heavily relies on structured knowledge. Named Entity Recognition (NER), an essential technique for extracting structured knowledge from text, has garnered significant research interest. However, current NER approaches face difficulties in handling multiple levels of entity nesting in knowledge extraction, especially within the manufacturing context. To address this issue, we present an Anchor Span-based NER (ASNER) approach for accurately and efficiently extracting manufacturing text knowledge. This method combines token and span classification. Initially, the head and tail features of the entity are extracted, with the corresponding anchor span of potential entities generated based on a boundary match neural network to localize the entities. Subsequently, token spatial contextual features are extracted using biaffine attention and convolutional neural networks, and classification categories are assigned to the anchor span-filtered entity features. Experimental studies demonstrate the effectiveness of the proposed ASNER approach. © 2024, The Authors. All rights reserved.

关键词： Character recognition

来源：评论

学校读者我要写书评

暂无评论

Global Consensus of Positive Edge System With Sector Input Nonlinearities

引用

IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS 2021年第7期51卷 4057-4066页

作者： Wu, Han Lam, James Su, Housheng Huazhong Univ Sci & Technol Sch Artificial Intelligence & Automat Image Proc & Intelligent Control Key Lab Educ Minist China Wuhan 430074 Peoples R China Wuhan Univ Sci & Technol Engn Res Ctr Met Automat & Measurement Technol Sch Informat Sci & Engn Wuhan 430081 Peoples R China Univ Hong Kong Dept Mech Engn Hong Kong Peoples R China

This paper investigates the positive edge consensus problems for undirected networks, where each edge is modeled by a continuous-time positive system with sector input nonlinearities. Based on a given nodal network, the corresponding edge network is presented, and some relationships between the given nodal network and its edge network are given. The observer-based protocol for non-negative consensus is proposed. Rigorous theoretical analysis which considers both the input nonlinearities and the positivity constraint is introduced to achieve a global consensus result. Simultaneous stabilization problems are constructed for edge-based non-negative consensus using the maximum and the minimum nonzero eigenvalues of the edge Laplacian matrix. Furthermore, the robustness of the positive edge consensus to parameter perturbation is illustrated. Sufficient non-negative consensus conditions related to the bounds of system matrices are given, and the non-negative edge consensus conditions for star nodal networks are also given only using the vertex number. Iterative linear matrix inequalities (LMIs)/LMI algorithm is presented to obtain the feedback matrix and the observer matrix. The given theoretical results are finally illustrated by some simulations.

关键词： Vehicle dynamics Linear matrix inequalities Observers Topology Complex networks Protocols control systems Edge network observer-based consensus positive system sector nonlinearity

来源：评论

学校读者我要写书评

暂无评论

Temporal Context Aggregation Network for Temporal Action Proposal Refinement

Temporal Context Aggregation Network for Temporal Action Pro...

引用

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Qing, Zhiwu Su, Haisheng Gan, Weihao Wang, Dongliang Wu, Wei Wang, Xiang Qiao, Yu Yan, Junjie Gao, Changxin Sang, Nong Huazhong Univ Sci & Technol Key Lab Image Proc & Intelligent Control Sch Artificial Intelligence & Automat Wuhan Peoples R China SenseTime Res Hong Kong Peoples R China Chinese Acad Sci Shenzhen Inst Adv Technol Beijing Peoples R China Shanghai AI Lab Shanghai Peoples R China

ISBN: (纸本)9781665445092

Temporal action proposal generation aims to estimate temporal intervals of actions in untrimmed videos, which is a challenging yet important task in the video understanding field. The proposals generated by current methods still suffer from inaccurate temporal boundaries and inferior confidence used for retrieval owing to the lack of efficient temporal modeling and effective boundary context utilization. In this paper, we propose Temporal Context Aggregation Network (TCANet) to generate high-quality action proposals through "local and global" temporal context aggregation and complementary as well as progressive boundary refinement. Specifically, we first design a Local-Global Temporal Encoder (LGTE), which adopts the channel grouping strategy to efficiently encode both "local and global" temporal inter-dependencies. Furthermore, both the boundary and internal context of proposals are adopted for frame-level and segment-level boundary regressions, respectively. Temporal Boundary Regressor (TBR) is designed to combine these two regression granularities in an end-to-end fashion, which achieves the precise boundaries and reliable confidence of proposals through progressive refinement. Extensive experiments are conducted on three challenging datasets: HACS, ActivityNet-v1.3, and THUMOS-14, where TCANet can generate proposals with high precision and recall. By combining with the existing action classifier, TCANet can obtain remarkable temporal action detection performance compared with other methods. Not surprisingly, the proposed TCANet won the 1st place in the CVPR 2020 - HACS challenge leaderboard on temporal action localization task.

关键词： Location awareness Computer vision Benchmark testing Reliability engineering Pattern recognition Proposals Task analysis

来源：评论

学校读者我要写书评

暂无评论

Machine learning based soft sensor and long-term calibration scheme: A solid oxide fuel cell system case

引用

INTERNATIONAL JOURNAL OF HYDROGEN ENERGY 2021年第33期46卷 17322-17342页

作者： Zhao, Weiqi Jiang, Jianhua Qin, Hongchuan Li, Xi Li, Jian Huazhong Univ Sci & Technol Educ Minist Image Proc & Intelligent Control Key Lab Sch Artificial Intelligence & Automat Wuhan 430074 Hubei Peoples R China Huazhong Univ Sci & Technol China EU Inst Clean & Renewable Energy Wuhan 430074 Hubei Peoples R China Huazhong Univ Sci & Technol Sch Mat Sci & Engn Wuhan 430074 Hubei Peoples R China

To quickly and accurately obtain the multi-dimensional state in a sealing reactor is critical for the safety and high performance of industrial system. In this work, a generic machine learning based multi-dimensional soft sensor and a long-term calibration scheme is proposed by using the case of steam reforming solid oxide fuel cell (SR-SOFC) system. Firstly, based on a validated SR-SOFC system model, the temporal-spatial temperature distribution (TSTD) characteristics are analyzed and a TSTD characteristic model is constructed by Multivariable Linear Regression. Then, the central node temperature of the stack is estimated by applying the Least Square Support Vector Machine and the SR-SOFC stack temperature distribution is accurately obtained by explaining the TSTD characteristics model with the central node temperature. In addition, the temperature distribution is calibrated by Stochastic Gradient Descent algorithm to eliminate the estimation error resulting from stack degradation in a long term. The simulation results show that the proposed method can obtain the SR-SOFC stack temperature distribution in time and effectively, the average error is less than 1 K. The proposed estimation strategy can be easily expanded to other scenarios with similar operation conditions. (c) 2021 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

关键词： Soft sensor Solid oxide fuel cell system Temperature distribution Pattern recognition Temporal-spatial characteristics Stack degradation

来源：评论

学校读者我要写书评

暂无评论

Understanding the Robustness of 3D Object Detection with Bird'View Representations in Autonomous Driving

Understanding the Robustness of 3D Object Detection with Bir...

引用

Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Zijian Zhu Yichi Zhang Hai Chen Yinpeng Dong Shu Zhao Wenbo Ding Jiachen Zhong Shibao Zheng Institute of Image Communication and Network Engineering Shanghai Jiao Tong University Dept. of Comp. Sci. and Tech. THBI Lab Institute for AI Tsinghua University BNRist Center Zhongguancun Laboratory Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Information Materials and Intelligent Sensing Laboratory of Anhui Province School of Computer Science and Technology Anhui University SAIC Motor AI Lab

3D object detection is an essential perception task in autonomous driving to understand the environments. The Bird's-Eye-View (BEV) representations have significantly improved the performance of 3D detectors with camera inputs on popular benchmarks. However, there still lacks a systematic understanding of the robustness of these vision-dependent BEV models, which is closely related to the safety of autonomous driving systems. In this paper, we evaluate the natural and adversarial robustness of various representative models under extensive settings, to fully understand their behaviors influenced by explicit BEV features compared with those without BEV. In addition to the classic settings, we propose a 3D consistent patch attack by applying adversarial patches in the 3D space to guarantee the spatiotemporal consistency, which is more realistic for the scenario of autonomous driving. With substantial experiments, we draw several findings: 1) BEV models tend to be more stable than previous methods under different natural conditions and common corruptions due to the expressive spatial representations; 2) BEV models are more vulnerable to adversarial noises, mainly caused by the redundant BEV features; 3) Camera-LiDARfusion models have superior performance under different settings with multi-modal inputs, but BEV fusion model is still vulnerable to adversarial noises of both point cloud and image. These findings alert the safety issue in the applications of BEV detectors and could facilitate the development of more robust models.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A memristive dual-slope A/D converter

引用

INTERNATIONAL JOURNAL OF CIRCUIT THEORY AND APPLICATIONS 2020年第1期48卷 42-55页

作者： Yang, Le Zeng, Zhigang Ma, Zhansong Shan, Wenqi Huazhong Univ Sci & Technol Sch Artificial Intelligence & Automat Wuhan 430074 Hubei Peoples R China Huazhong Univ Sci & Technol Key Lab Image Proc & Intelligent Control Educ Minist China Wuhan Hubei Peoples R China

As an emerging device, memristor has several excellent properties like changeable memristance, nonvolatility, and nanoscale. Based on complementary metal-oxide-semiconductor (CMOS) dual-slope analog-to-digital (A/D) converter, this paper proposes a memristive dual-slope A/D converter. Owing to the usage of memristor, the proposed memristive A/D converter not only has more compact circuit structure and simpler control timing than the CMOS one but also has advantages over the existing memristive conversion circuits in circuit design and application. For the memristive A/D converter, a conversion process consists of two count procedures. By means of controlling the memristance change in the two count procedures, the A/D converter converts an analog signal to the corresponding digital count value. Meanwhile, the conversion result is inferred according to the circuit structure of the A/D converter. Then, combining the conversion process and PSPICE simulation, this paper analyzes the anti-interference performance of the A/D converter. Further, the robustness of the A/D converter is presented, applying the similar analysis methods. The analysis results demonstrate that the proposed A/D converter has good anti-interference and robustness performances.

关键词： anti-interference dual-slope A D converter memristive A D converter memristor robustness

来源：评论

学校读者我要写书评

暂无评论

Output Multiformation Tracking of Networked Heterogeneous Robotic Systems via Finite-Time Hierarchical control

引用

IEEE TRANSACTIONS ON CYBERNETICS 2021年第6期51卷 2893-2904页

作者： Liang, Chang-Duo Ge, Ming-Feng Liu, Zhi-Wei Wang, Yan-Wu Karimi, Hamid Reza China Univ Geosci Sch Mech Engn & Elect Informat Wuhan 430074 Peoples R China 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 430074 Peoples R China Politecn Milan Dept Mech Engn I-20156 Milan Italy

This article investigates the finite-time output multiformation tracking (OMFT) problem of networked heterogeneous robotic systems (NHRSs), where each robot model involves external disturbances, parametric uncertainties, and possible kinematic redundancy. Besides, the interactions among robotic systems are described as a directed graph with an acyclic partition. Then, several novel practical finite-time hierarchical control (FTHC) algorithms are designed. The convergence analysis of the closed-loop dynamics is extremely difficult due to the lack of effective analysis methods. Based on the mathematics induction and reductio ad absurdum, a new nonsmooth Lyapunov function is proposed to derive the sufficient conditions and settling time functions. Finally, numerical simulations are performed on the NHRS to verify the main results.

关键词： Robot kinematics Convergence Directed graphs Multi-agent systems Task analysis Kinematics Finite-time hierarchical control (FTHC) networked heterogeneous robotic systems (NHRSs) nonsmooth Lyapunov function output multiformation tracking (OMFT)

来源：评论

学校读者我要写书评

暂无评论

Resilient control and Analysis for DC Microgrid System Under DoS and Impulsive FDI Attacks

引用

IEEE TRANSACTIONS ON SMART GRID 2021年第5期12卷 3742-3754页

作者： Liu, Xiao-Kang Wen, Changyun Xu, Qianwen Wang, Yan-Wu Huazhong Univ Sci & Technol Sch Artificial Intelligence & Automat Wuhan 430074 Peoples R China Huazhong Univ Sci & Technol Minist Educ Key Lab Image Proc & Intelligent Control Wuhan 430074 Peoples R China Nanyang Technol Univ Sch Elect & Elect Engn Singapore 639798 Singapore KTH Royal Inst Technol Elect Power & Energy Syst Div S-10044 Stockholm Sweden

This paper proposes a resilient controller for DC microgrid to achieve current sharing and voltage restoration under discrete-time false data injection (FDI) and denial-of-service (DoS) attacks. Switching and impulsive signals are used to model the dynamic system of DC microgrid under DoS and FDI. To deal with the cyber attacks, a combined error of current and voltage is proposed and a switching secondary controller is designed. Based on the stability analysis method on hybrid systems, we establish a sufficient condition for selecting control parameters in relation to the average dwell time of FDI attack and the normal communication rate under DoS attack. Furthermore, an adaptive gain based control scheme is proposed to relax the requirement on knowledge of the cyber attacks in control parameter design. The utility of the results is illustrated through case studies on a tested DC microgrid.

关键词： Microgrids Voltage control Cyberattack Stability analysis Switches Matrix converters Resistance DC microgrid consensus control stability analysis cyber-attacks

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：