检索结果-内蒙古大学图书馆

Dynamic Privacy-preserving Collaborative Schemes for Average Computation

IFAC-PapersOnLine 2020年第2期53卷 2963-2968页

作者： Xin Wang Hideaki Ishii Jianping He Peng Cheng State Key Lab. of Industrial Control Technology Zhejiang University Hangzhou 310027 China Dept. of Computer Science Tokyo Institute of Technology Yokohama 226-8502 Japan Dept. of Automation Shanghai Jiao Tong University Shanghai 200240 China

In this paper, we consider the privacy-preserving problem in collaborative computing. Based on a two-step average computation framework, we propose three privacy-aware schemes, all of which achieve different levels of privacy protections depending on data servers’ trust degrees. Further, by carefully designing noises injected to the distributed computing process, we obtain dynamic privacy-preserving schemes, whose privacy preserving levels are measured by Kullback-Leibler differential privacy. In addition, we prove that the proposed schemes achieve convergence in different senses. Numerical experiments are finally conducted to verify the obtained privacy properties and convergence guarantees.

关键词： Collaborative computing dynamic privacy average consensus convergence

来源：评论

学校读者我要写书评

暂无评论

Target-enclosing affine formation control of two-layer networked spacecraft with collision avoidance

引用

Chinese Journal of Aeronautics 2019年第12期32卷 2679-2693页

作者： Yang XU Delin LUO Dongyu LI Yancheng YOU Haibin DUAN School of Aerospace Engineering Xiamen UniversityXiamen 361005China Department of Control Science and Engineering Harbin Institute of TechnologyHarbin 150001China State Key Laboratory of Virtual Reality Technology and Systems School of Automation Science and Electrical EngineeringBeihang UniversityBeijing 100083China Department of Electrical&Computer Engineering National University of SingaporeSingapore 117583Singapore

This paper addresses a target-enclosing problem for multiple spacecraft systems by proposing a two-layer affine formation control strategy. Compared with the existing methods,the adopted two-layer network structure in this paper is generally directed, which is suitable for practical space missions. Firstly, distributed finite-time sliding-mode estimators and formation controllers in both layers are designed separately to improve the flexibility of the formation control system. By introducing the properties of affine transformation into formation control protocol design,the controllers can be used to track different time-varying target formation patterns. Besides, multilayer time-varying encirclements can be achieved with particular shapes to surround the moving target. In the sequel, by integrating adaptive neural networks and specialized artificial potential functions into backstepping controllers, the problems of uncertain Euler-Lagrange models, collision avoidance as well as formation reconfiguration are solved simultaneously. The stability of the proposed controllers is verified by the Lyapunov direct method. Finally, two simulation examples of triangle formation and more complex hexagon formation are presented to illustrate the feasibility of the theoretical results.

关键词： Affine formation control Collision avoidance Lyapunov stability Target enclosing Two-layer strategy

来源：评论

学校读者我要写书评

暂无评论

Deep Rank-Consistent Pyramid Model for Enhanced Crowd Counting

arXiv

引用

arXiv 2022年

作者： Gao, Jiaqi Huang, Zhizhong Lei, Yiming Shan, Hongming Wang, James Z. Wang, Fei-Yue Zhang, Junping The Shanghai Key Laboratory of Intelligent Information Processing School of Computer Science Fudan University Shanghai200433 China The Institute of Science and Technology for Brain-inspired Intelligence MOE Frontiers Center for Brain Science Fudan University Shanghai200433 China The Shanghai Center for Brain Science and Brain-Inspired Technology Shanghai201210 China The College of Information Sciences and Technology The Pennsylvania State University University ParkPA16802 United States The State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences Beijing100190 China The Institute of Systems Engineering Macau University of Science and Technology China Qingdao Academy of Intelligent Industries Qingdao266109 China

Most conventional crowd counting methods utilize a fully-supervised learning framework to establish a mapping between scene images and crowd density maps. They usually rely on a large quantity of costly and time-intensive pixel-level annotations for training supervision. One way to mitigate the intensive labeling effort and improve counting accuracy is to leverage large amounts of unlabeled images. This is attributed to the inherent self-structural information and rank consistency within a single image, offering additional qualitative relation supervision during training. Contrary to earlier methods that utilized the rank relations at the original image level, we explore such rank-consistency relation within the latent feature spaces. This approach enables the incorporation of numerous pyramid partial orders, strengthening the model representation capability. A notable advantage is that it can also increase the utilization ratio of unlabeled samples. Specifically, we propose a Deep Rank-consistEnt pyrAmid Model (DREAM), which makes full use of rank consistency across coarse-to-fine pyramid features in latent spaces for enhanced crowd counting with massive unlabeled images. In addition, we have collected a new unlabeled crowd counting dataset, FUDAN-UCC, comprising 4,000 images for training purposes. Extensive experiments on four benchmark datasets, namely UCF-QNRF, ShanghaiTech PartA and PartB, and UCF-CC-50, show the effectiveness of our method compared with previous semi-supervised methods. The codes are available at https://***/bridgeqiqi/DREAM. Copyright © 2022, 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

Stacked BNAS: Rethinking Broad Convolutional Neural Network for Neural Architecture Search

arXiv

引用

arXiv 2021年

作者： Ding, Zixiang Chen, Yaran Li, Nannan Zhao, Dongbin Chen, C.L. Philip State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences Beijing100190 China School of Artificial Intelligence University of Chinese Academy of Sciences Beijing100049 China School of Computer Science & Engineering South China University of Technology Guangdong Guangzhou510006 China College of Navigation Dalian Maritime University Dalian116026 China

Different from other deep scalable architecture-based NAS approaches, Broad Neural Architecture Search (BNAS) proposes a broad scalable architecture which consists of convolution and enhancement blocks, dubbed Broad Convolutional Neural Network (BCNN), as the search space for amazing efficiency improvement. BCNN reuses the topologies of cells in the convolution block so that BNAS can employ few cells for efficient search. Moreover, multi-scale feature fusion and knowledge embedding are proposed to improve the performance of BCNN with shallow topology. However, BNAS suffers some drawbacks: 1) insufficient representation diversity for feature fusion and enhancement and 2) time consumption of knowledge embedding design by human experts. This paper proposes Stacked BNAS, whose search space is a developed broad scalable architecture named Stacked BCNN, with better performance than BNAS. On the one hand, Stacked BCNN treats mini BCNN as a basic block to preserve comprehensive representation and deliver powerful feature extraction ability. For multi-scale feature enhancement, each mini BCNN feeds the outputs of deep and broad cells to the enhancement cell. For multi-scale feature fusion, each mini BCNN feeds the outputs of deep, broad and enhancement cells to the output node. On the other hand, Knowledge Embedding Search (KES) is proposed to learn appropriate knowledge embeddings in a differentiable way. Moreover, the basic unit of KES is an over-parameterized knowledge embedding module that consists of all possible candidate knowledge embeddings. Experimental results show that 1) Stacked BNAS obtains better performance than BNAS-v2 on both CIFAR-10 and ImageNet, 2) the proposed KES algorithm contributes to reducing the parameters of the learned architecture with satisfactory performance, and 3) Stacked BNAS delivers a state-of-the-art efficiency of 0.02 GPU days. Copyright © 2021, The Authors. All rights reserved.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

CNN-based Broad Learning with Efficient Incremental Reconstruction Model for Facial Emotion Recognition ⁎

引用

IFAC-PapersOnLine 2020年第2期53卷 10236-10241页

作者： Luefeng Chen Min Li Xuzhi Lai Kaoru Hirota Witold Pedrycz School of Automation China University of Geosciences Wuhan 430074 China Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Wuhan 430074 China Tokyo Institute of Technology Yokohama 226-8502 Japan Department of Electrical and Computer Engineering University of Alberta Edmonton AB T6R 2G7 Canada

Convolutional neural network-based broad learning with efficient incremental reconstruction model (CNNBL) is proposed to recognize emotions in human-robot interaction. It aims to extract deep and abstract features from facial emotional images, and reduce the influence of the complex structure and slow network updates on facial emotion recognition in deep learning. Feature extraction is carried out by convolution and maximum pooling, and then the ridge regression algorithm is used for emotion recognition. When the network needs to expand, the network is dynamically updated by incremental learning algorithm. We verified the experimental performance through k -fold cross validation. According to the recognition results, the accuracy on JAFFE database of our proposal is greater than that of the state of the art, such as the Local Binary Patterns with Softmax and Deep Attentive Multi-path convolutional neural network.

关键词： Convolution Neural Networks Broad Learning Emotion Recognition Human-Robot Interaction

来源：评论

学校读者我要写书评

暂无评论

Measurement and Calculation of the Impedance of an Eddy Current Probe Placed Above a Disc with Two Layers of Different Diameters

引用

Materials (Basel, Switzerland) 2025年第10期18卷 2376页

作者： Yike Xiang Grzegorz Tytko Yao Luo Jolanta Makowska State Key Laboratory of Power Grid Environmental Protection School of Electrical Engineering and Automation Wuhan University Wuhan 430072 China. School of Electrical Engineering and Automation Wuhan University Wuhan 430072 China. Faculty of Automatic Control Electronics and Computer Science The Silesian University of Technology Akademicka 16 44-100 Gliwice Poland. Institute of Materials Engineering Faculty of Science and Technology University of Silesia 75 Pułku Piechoty 1A 41-500 Chorzow Poland.

This work presents a system developed to determine changes in the impedance of an eddy current probe placed above a conductive disc containing two layers of different diameters. In the first step, an analytical model was derived with an employment of the truncated region eigenfunction expansion (TREE) method. The final formula for the probe impedance change was presented in a closed form, which makes it possible to implement it in any programming language or computer algebra system. The mathematical model was implemented in MATLAB and used to design probes and to determine the optimal test parameters. In the next step, two eddy current probes with a single coil with different geometric dimensions were constructed. Impedance measurements were carried out using an LCR meter for three sets of double-layer discs. The tested discs were made of materials with different electrical conductivities. The upper and lower layers of the disc also differed in terms of the geometric dimensions, i.e., the diameter and thickness. The tests were performed for the operating frequency of the probe ranging from 1 kHz to 10 kHz. In all cases, a very good agreement was obtained between the measurement and the calculation results. Both the error in the changes in resistance and the error in the changes in reactance did not exceed 3.5%.

关键词： analytical model coil reactance coil resistance conductive disc eddy current testing probe impedance truncated region eigenfunction expansion (TREE)

来源：评论

学校读者我要写书评

暂无评论

Networked Event-Based Trajectory Planning for Mobile Objects

Networked Event-Based Trajectory Planning for Mobile Objects

引用

作者： Schwung, Michael Lunze, Jan Institute of Automation and Computer Control Ruhr-Universität Bochum Germany

This paper proposes an event-based method to ensure a collision-free movement of mobile objects. The objects are locally controlled and connected by a communication network. Each object possesses an event-based trajectory planning unit, which acts on a dynamical model representing the neighbouring objects together with the uncertainty of the predicted states. If a common control goal is violated due to local disturbances affecting the objects, communication to neighbouring objects is invoked. A replanning of the local trajectory based on the communicated information attenuates the disturbance. In contrast to existing literature, this approach has a lower communication effort. The effectiveness of the method is illustrated by a simulation study with two quadrotors. © 2018

关键词： Trajectories

来源：评论

学校读者我要写书评

暂无评论

Tracking control of Networked and Interconnected Systems

Tracking Control of Networked and Interconnected Systems

引用

作者： Schenk, Kai Lunze, Jan Institute of Automation and Computer Control Ruhr-Universität Bochum Germany

Networked and interconnected systems denote a class of systems that consists of physically coupled subsystems that can exchange information via a communication network. This paper presents a new design method for tracking controllers, which guarantee that the output of each subsystem follows a specified trajectory. Each controller is synthesised by a design agent that is introduced as a local component with limited knowledge of the overall system. Each design agent uses the network to gain information that is necessary for the controller synthesis. The information exchange takes places exclusively at design time and concerns an important aspect of this work. The method presented in this paper extends classical two-degree-of-freedom controllers by decoupling units. These units are essential for achieving the desired tracking. It is proven that the proposed extension yields the desired trajectory tracking. A simulation study demonstrates the applicability. © 2018

关键词： Trajectories

来源：评论

学校读者我要写书评

暂无评论

Regret and Cumulative Constraint Violation Analysis for Distributed Online Constrained Convex Optimization

arXiv

引用

arXiv 2021年

作者： Yi, Xinlei Li, Xiuxian Yang, Tao Xie, Lihua Chai, Tianyou Johansson, Karl H. The Division of Decision and Control Systems School of Electrical Engineering and Computer Science KTH Royal Institute of Technology Stockholm10044 Sweden The Digital Futures Stockholm10044 Sweden The Department of Control Science and Engineering College of Electronics and Information Engineering The Shanghai Research Institute for Intelligent Autonomous Systems Tongji University Shanghai China The State Key Laboratory of Synthetical Automation for Process Industries Northeastern University Shenyang110819 China School of Electrical and Electronic Engineering Nanyang Technological University 50 Nanyang Avenue Singapore639798 Singapore

This paper considers the distributed online convex optimization problem with time-varying constraints over a network of agents. This is a sequential decision making problem with two sequences of arbitrarily varying convex loss and constraint functions. At each round, each agent selects a decision from the decision set, and then only a portion of the loss function and a coordinate block of the constraint function at this round are privately revealed to this agent. The goal of the network is to minimize the network-wide loss accumulated over time. Two distributed online algorithms with full-information and bandit feedback are proposed. Both dynamic and static network regret bounds are analyzed for the proposed algorithms, and network cumulative constraint violation is used to measure constraint violation, which excludes the situation that strictly feasible constraints can compensate the effects of violated constraints. In particular, we show that the proposed algorithms achieve O(T max{κ,1-κ}) static network regret and O(T 1-κ/2) network cumulative constraint violation, where T is the time horizon and κ ∈ (0, 1) is a user-defined trade-off parameter. Moreover, if the loss functions are strongly convex, then the static network regret bound can be reduced to O(T κ). Finally, numerical simulations are provided to illustrate the effectiveness of the theoretical results. Copyright © 2021, The Authors. All rights reserved.

关键词： Economic and social effects

来源：评论

学校读者我要写书评

暂无评论

Multi-hop reasoning over sparse knowledge graphs with deep reinforcement learning

引用

Expert Systems with Applications 2025年 290卷

作者： Yiyang Liu Lejun Fu Yukai Fu Tianxing Wu Hongfei Bai State Key Laboratory of Robotics Shenyang Institute of Automation Chinese Academy of Sciences Shenyang 110016 PR China Key Laboratory of Networked Control Systems Chinese Academy of Sciences Shenyang 110016 PR China College of Information Engineering Shenyang University of Chemical Technology Shenyang 110142 PR China School of Computer Science and Engineering Southeast University Nanjing 211189 PR China

In recent years, knowledge reasoning has demonstrated substantial progress in inferring missing information, effectively alleviating the limitations caused by the incompleteness of knowledge graphs (KGs). Dealing with entities that have sufficient paths between them, multi-hop reasoning methods have achieved notable success and demonstrated interpretability, but they still exhibit deficiencies when dealing with sparse KGs. To overcome the insufficiency of edges in sparse KGs and the absence of effective guidance in multi-hop reasoning methods, a hybrid multi-hop collaborative Multi-Agent Sparse Knowledge Graph Reasoning model (MASKGR) is proposed. Attention-based dynamic completion (ADC) and variable rule guidance(VRG) are used to enhance the reasoning capability of deep reinforcement learning (RL) in sparse KGs. ADC focuses on high-value relations based on attention weights and combines embedding-based methods to predict the tail entities of these relations, expanding the original action space and improving the exploration capability of agents. Moreover, VRG greatly improves the ability of MASKGR to use rule-guided agent learning by utilizing the variable rule ratio and the variable loss reshaping. ADC can be applied to both single-agent and multi-agent in other models, while VRG is specifically designed for multi-agent. Comprehensive experiments have been conducted on four datasets: FB15K-237-10 %, FB15K-237-50 %, NELL23K, and WD-singer. Specifically, MASKGR achieves relative improvements of 1.3 % in Mean Reciprocal Rank (MRR) and 1.1 % in the proportion of correct tail entity rankings in the top 3 (Hit@3) on the FB15K-237-10 %, 2.4 % in MRR and 1.3 % in Hit@3 on the NELL23K, and 2.7 % in MRR and 3.6 % in Hit@3 on the WD-singer. The results indicate that MASKGR outperforms state-of-the-art multi-hop reasoning models based on reinforcement learning.

关键词： Deep reinforcement 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：