检索结果-内蒙古大学图书馆

Combining particle filter algorithm with bio-inspired anemotaxis behavior: A smoke plume tracking method and its robotic experiment validation

引用

MEASUREMENT 2020年第0期154卷 107482-000页

作者： Chen, Xinxing Huang, Jian Huazhong Univ Sci & Technol Sch Artificial Intelligence & Automat Key Lab Minist Educ Image Proc & Intelligent Control Wuhan Peoples R China

The smoke source localization robot is a safe substitution of human and animal rescuers in many dangerous search and rescue mission. In this paper, we propose an anemotaxis - particle filter based smoke plume path tracking method and presented a smoke source localization robot. A modified firefly algorithm is applied in the resampling step of particle filter based smoke plume tracking method to mimic anemotaxis behaviors in the odor source searching process of living creatures. The performance of the proposed algorithm is qualitatively evaluated in a simulated wind tunnel environment in Webots, a high-fidelity robotic simulator. The average computation time including plume path estimation, resampling and target evaluation in every step is 35.9974 ms and the average smoke source localization error is 0.7769 m. We also verify the feasibility by running the proposed algorithm on the smoke source localization robot in real experiments. (C) 2020 Elsevier Ltd. All rights reserved.

关键词： Olfactory robot Particulate matter Firefly algorithm Particle filter Entropy

来源：评论

学校读者我要写书评

暂无评论

Dual Attention Mechanism Based Outline Loss for image Stylization

引用

NEURAL processing LETTERS 2023年第1期55卷 537-549页

作者： Tu, Pengqi Sang, Nong Huazhong Univ Sci & Technol Sch Artificial Intelligence & Automat Minist Educ Image Proc & Intelligent Control Key Lab Wuhan Peoples R China

image stylization has attracted considerable attention from various fields. Although impressive results have been achieved, existing methods pay less attention to the preservation of outline and putting constraint on it when training, which makes generated images suffering from different degrees of distortion. To address this issue, we propose a dual attention mechanism based outline loss to enhance the restriction of outline consistency by incorporating an outline detection module and a dual attention module. Specifically, an outline detection module is used to detect outlines of the source image and the stylized image, which are further compared and enforced to be consistent with each other by a carefully-elaborated outline loss. Additionally, the dual attention module first guides the model to focus on regions of the source image whose style has the biggest difference from the target image during stylization based on the style attention feature map obtained by the auxiliary classifier. Then, an outline attention map is predicted to highlight regions where the outlines are prone to distort during stylization, which further facilitates the outline loss to execute stronger constraint on these regions. Experimental results show the superiority of our method compared to the existing state-of-the-art methods

关键词： image stylization Dual attention mechanism Outline loss

来源：评论

学校读者我要写书评

暂无评论

Enhancing Deep Learning-based Depression Level Estimation Based on Multi-task Learning 24

Enhancing Deep Learning-based Depression Level Estimation Ba...

引用

7th International Conference on Machine Learning and Machine Intelligence, MLMI 2024

作者： Ye, Qingxin Xie, Zhenming Sun, Hao Xin, Luyao Chen, Youwen Song, Jian Chen, Yen-Wei Provincial key Laboratory of Computational Science Huaqiao University Quanzhou China Intelligent Image Processing Lab Zhejiang University Hangzhou China Intelligent Image Processing Lab Ritsumeikan University Osaka Japan

ISBN: (纸本)9798400717833

In recent years, depression, as a serious mental illness, has received widespread attention from various sectors of society. How to identify depressive emotions in a timely manner and detect depression has become an urgent issue that needs to be addressed. Currently, many experts and scholars propose using deep learning methods to assess the level of depression. Compared to other deep learning models, Transformer models have parallel computing capabilities with their self-attention mechanism, enabling them to handle entire sequences and capture long-distance dependencies. After the success of Transformers in natural language processing, similar architectures have been introduced into the field of computer vision, promoting advancements in multi-modal research. However, how to integrate other information or tasks to improve assessment accuracy remains a challenge. In our team's preliminary work, we proposed a multi-modal adaptive fusion Transformer network for assessing depression levels. Inspired by multi-task learning, we introduced gender classification as a second auxiliary task alongside the main task of depression regression, forming a three-task learning framework including depression classification. This was done to enhance the model's performance. We validated the effectiveness of the proposed method by testing it on a public dataset (AVEC 2019 Detecting Depression with AI Sub-challenge). The experimental results show that our proposed method achieved a Concordance Correlation Coefficient (CCC) of 0.60448, which is 1.688% higher than the accuracy of the dual-task learning method (CCC = 0.5876). © 2024 Copyright held by the owner/author(s).

关键词： Multi-task learning

来源：评论

学校读者我要写书评

暂无评论

Generalized Accurate Harmonic Calculation Method Based on Discretized Double Fourier Series to Solve Double-Pulse Phenomenon

引用

IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS 2023年第6期70卷 5651-5661页

作者： Huang, Songtao Xu, Jinbang Ye, Jie Shen, Anwen Huazhong Univ Sci & Technol Sch Artificial Intelligence & Automat Natl Key Lab Sci & Technol Multispectral Informat Key Lab Image Proc & Intelligent Control Wuhan 430074 Peoples R China

Harmonic issues have always been paid attention to in electrical systems, and accurate calculation of the amplitude and phase of each harmonic has become one of the research focuses of scholars. This article develops a generalized harmonic closed-form analytical solution for a single-phase voltage source inverter under digital sinusoidal pulsewidth modulation and dead-time effects. This article is the first to point out the double-pulse phenomenon that can occur when using a double Fourier series to solve the harmonic spectrum due to the discontinuity of the dead-time effects and half-cycle integrals. Analysis and data comparison show that this phenomenon can affect the accuracy of harmonic calculations. A discretized double Fourier series method is proposed to solve this phenomenon. Moreover, the problems of missing pulse caused by the modulation under a specific current phase and current zero-crossing are also solved, improving the harmonic calculation's accuracy and generality. Simulation analysis and the results of inverse fast Fourier transform proved the correctness of the optimization method, and the experimental results are obtained on the DSP and SIC platforms, which are in good agreement with the expression calculation results.

关键词： Harmonic analysis Power system harmonics Manganese Fourier series Voltage Analytical models Phase modulation Dead-time effects digital sinusoidal pulse width modulation (DSPWM) discretized double Fourier series double-pulse phenomenon harmonic calculation voltage source inverter (VSI)

来源：评论

学校读者我要写书评

暂无评论

Passivity-Based Robust control Against Quantified False Data Injection Attacks in Cyber-Physical Systems

引用

IEEE/CAA Journal of Automatica Sinica 2021年第8期8卷 1440-1450页

作者： Yue Zhao Ze Chen Chunjie Zhou Yu-Chu Tian Yuanqing Qin Key Laboratory of Ministry of Education for Image Processing and Intelligent Control School of Artificial Intelligence and AutomationHuazhong University of Science and TechnologyWuhan 430074China School of Computer Science Queensland University of TechnologyBrisbane QLD 4001Australia

Secure control against cyber attacks becomes increasingly significant in cyber-physical systems(CPSs).False data injection attacks are a class of cyber attacks that aim to compromise CPS functions by injecting false data such as sensor measurements and control *** quantified false data injection attacks,this paper establishes an effective defense framework from the energy conversion ***,we design an energy controller to dynamically adjust the system energy changes caused by unknown *** designed energy controller stabilizes the attacked CPSs and ensures the dynamic performance of the system by adjusting the amount of damping ***,with the disturbance attenuation technique,the burden of control system design is simplified because there is no need to design an attack *** addition,this secure control method is simple to implement because it avoids complicated mathematical *** effectiveness of our control method is demonstrated through an industrial CPS that controls a permanent magnet synchronous motor.

关键词： Cyber-physical systems energy controller energy conversion false data injection attacks L2 disturbance attenuation technology

来源：评论

学校读者我要写书评

暂无评论

CondNet: Conditional Classifier for Scene Segmentation

引用

IEEE SIGNAL processing LETTERS 2021年 28卷 758-762页

作者： Yu, Changqian Shao, Yuanjie Gao, Changxin Sang, Nong Huazhong Univ Sci & Technol Sch Artificial Intelligence & Automat Key Lab Image Proc & Intelligent Control Wuhan 430074 Peoples R China

The fully convolutional network (FCN) has achieved tremendous success in dense visual recognition tasks, such as scene segmentation. The last layer of FCN is typically a global classifier (1 x 1 convolution) to recognize each pixel to a semantic label. We empirically show that this global classifier, ignoring the intra-class distinction, may lead to sub-optimal results. In this work, we present a conditional classifier to replace the traditional global classifier, where the kernels of the classifier are generated dynamically conditioned on the input. The main advantages of the new classifier consist of: (i) it attends on the intra-class distinction, leading to stronger dense recognition capability;(ii) the conditional classifier is simple and flexible to be integrated into almost arbitrary FCN architectures to improve the prediction. Extensive experiments demonstrate that the proposed classifier performs favourably against the traditional classifier on the FCN architecture. The framework equipped with the conditional classifier (called CondNet) achieves new state-of-the-art performances on two datasets. The code and models are available at https://***/CondNet.

关键词： Kernel Convolution Feature extraction Semantics Training Task analysis Aggregates Conditional classifier dynamic convolutions semantic segmentation

来源：评论

学校读者我要写书评

暂无评论

An improved estimation of distribution algorithm for multi-compartment electric vehicle routing problem

引用

Journal of Systems Engineering and Electronics 2021年第2期32卷 365-379页

作者： SHEN Yindong PENG Liwen LI Jingpeng Key Laboratory of Image Processing and Intelligent Control School of Artificial Intelligence and AutomationHuazhong University of Science and TechnologyWuhan 430074China Division of Computing Science and Mathematics University of StirlingStirling FK94LAUK

The multi-compartment electric vehicle routing problem(EVRP)with soft time window and multiple charging types(MCEVRP-STW&MCT)is studied,in which electric multi-compartment vehicles that are environmentally friendly but need to be recharged in course of transport process,are employed.A mathematical model for this optimization problem is established with the objective of minimizing the function composed of vehicle cost,distribution cost,time window penalty cost and charging service *** solve the problem,an estimation of the distribution algorithm based on Lévy flight(EDA-LF)is proposed to perform a local search at each iteration to prevent the algorithm from falling into local *** results demonstrate that the EDA-LF algorithm can find better solutions and has stronger robustness than the basic EDA *** addition,when comparing with existing algorithms,the result shows that the EDA-LF can often get better solutions in a relatively short time when solving medium and large-scale *** experiments show that using electric multi-compartment vehicles to deliver incompatible products can produce better results than using traditional fuel vehicles.

关键词： multi-compartment vehicle routing problem electric vehicle routing problem(EVRP) soft time window multiple charging type estimation of distribution algorithm(EDA) Lévy flight

来源：评论

学校读者我要写书评

暂无评论

Optimal ascent guidance based on direct optimization method

Optimal ascent guidance based on direct optimization method

引用

2012 International Conference on Modelling, Identification and control, ICMIC 2012

作者： Lu, Xuefang Wang, Yongji Key Laboratory of Ministry of Education for Image Processing and Intelligent Control Huazhong University of Science and Technology Wuhan China

ISBN: (纸本)9780956715715

In this paper, an optimal guidance algorithm is proposed for atmospheric ascent. The optimal guidance algorithm updates the reference trajectory to deal with the impact of disturbance by solving an optimal control problem on-board. For the strong nonlinear ascent dynamic, the optimal control problem is transformed into nonlinear programming problem by trajectory discretization. The direct optimization method is implemented for this nonlinear programming problem. Only a segment of the reference trajectory is updated by the optimal guidance algorithm. Due to the small amount of the discrete nodes and the initial guess solution which is close to the optimization solution, the direct optimization method is fast enough to generate a new reference trajectory. Numerical simulation for the Generic Hypersonic Vehicle model and scramjet engine is done for different cases of the aerodynamic coefficient bias. The results show the accuracy and the effectiveness of this optimal guidance algorithm. © 2012 Huazhong Univ of Sci & Tec.

关键词： Trajectories

来源：评论

学校读者我要写书评

暂无评论

Ascent trajectory planning on-board based on direct optimisation method

引用

INTERNATIONAL JOURNAL OF MODELLING IDENTIFICATION AND control 2013年第4期19卷 388-395页

作者： Lu, Xuefang Wang, Yongji Huazhong Univ Sci & Technol Dept Control Sci & Engn Key Lab Minist Educ Image Proc & Intelligent Control Wuhan Hubei Peoples R China

In this paper, an on-board trajectory planning algorithm is proposed for atmospheric ascent. To deal with the impact of disturbance, the on-board trajectory planning algorithm updates the reference trajectory by solving an optimal control problem on-board. Considering the strong non-linear aerodynamic, the optimal control problem is transformed into non-linear programming problem by trajectory discretisation. The direct optimisation method is implemented for this non-linear programming problem. Due to the small amount of discrete nodes and the initial guess solution which is close to the optimisation solution, the direct optimisation method is fast enough to generate a new reference trajectory in every guidance cycle. With different cases of the aerodynamic coefficient bias, numerical simulation for the generic hypersonic vehicle model and scramjet engine is done. The results show the accuracy and the effectiveness of the on-board trajectory planning algorithm.

关键词： trajectory planning ascent direct optimisation non-linear programming problem

来源：评论

学校读者我要写书评

暂无评论

A Survey on Negative Transfer

引用

IEEE/CAA Journal of Automatica Sinica 2023年第2期10卷 305-329页

作者： Wen Zhang Lingfei Deng Lei Zhang Dongrui Wu the Key Laboratory of the Ministry of Education for Image Processing and Intelligent Control School of Artificial Intelligence and AutomationHuazhong University of Science and TechnologyWuhan 430074China IEEE the School of Microelectronics and Communication Engineering Chongqing UniversityChongqing 400044China

Transfer learning(TL)utilizes data or knowledge from one or more source domains to facilitate learning in a target *** is particularly useful when the target domain has very few or no labeled data,due to annotation expense,privacy concerns,***,the effectiveness of TL is not always *** transfer(NT),i.e.,leveraging source domain data/knowledge undesirably reduces learning performance in the target domain,and has been a long-standing and challenging problem in *** approaches have been proposed in the literature to address this ***,there does not exist a systematic *** paper fills this gap,by first introducing the definition of NT and its causes,and reviewing over fifty representative approaches for overcoming NT,which fall into three categories:domain similarity estimation,safe transfer,and NT *** areas,including computer vision,bioinformatics,natural language processing,recommender systems,and robotics,that use NT mitigation strategies to facilitate positive transfers,are also ***,we give guidelines on NT task construction and baseline algorithms,benchmark existing TL and NT mitigation approaches on three NT-specific datasets,and point out challenges and future research *** ensure reproducibility,our code is publicized at https://***/chamwen/NT-Benchmark.

关键词： Domain adaptation domain similarity negative transfer positive transfer transfer learning

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：