检索结果-内蒙古大学图书馆

arXiv 2018年

作者： Li, Kang Wu, Jinting Zhao, Xiaoguang Tan, Min State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences University of Chinese Academy of Sciences Beijing China

In this paper, we present a real-time Human-Robot Interaction (HRI) system for a service robot based on 3D human activity recognition and human-mimicking decision mechanism. A three-layer Long-Short-Term Memory (LSTM) network is trained to recognize human activity. The input of this network is the 3D skeleton joints information captured by Kinect V1 and the output of this network is the class label of human activity. Moreover, the human-mimicking decision mechanism is also fused into the online test process of human activity recognition, which allows the robot to instinctively decide whether to interrupt the current task according to task priority. The robot can make appropriate responses based on aforementioned analytical results. The system framework is realized on the Robot Operating System (ROS). The real-life activity interaction between our service robot and the user was conducted to demonstrate the effectiveness of developed HRI system. Copyright © 2018, The Authors. All rights reserved.

关键词： Human robot interaction

来源：评论

学校读者我要写书评

暂无评论

A Preliminary Research of Prediction Markets Based on Blockchain Powered Smart Contracts

A Preliminary Research of Prediction Markets Based on Blockc...

引用

IEEE/ACM Int'l Conference on & Int'l Conference on Cyber, Physical and Social Computing (CPSCom) Green Computing and Communications (GreenCom)

作者： Shuai Wang Xiaochun Ni Yong Yuan Fei-Yue Wang Xiao Wang Liwei Ouyang University of Chinese Academy of Sciences Beijing China The State Key Laboratory for Management and Control of Complex Systems Chinese Academy of Sciences Beijing China

Prediction markets are markets where participants trade contracts whose payoffs are tied to a future event, thereby yielding prices that can be interpreted as market aggregated forecasts. Past studies have shown that the prediction markets can provide accurate forecasts, sometimes better than sophisticated statistical tools. Due to their advantages, prediction markets have been widely used in the prediction of elections, project management, product quality, and impact of events. However, prediction markets also have some limitations, e.g., poor anonymity and limited market liquidity. In this paper, we propose to apply blockchain powered smart contracts to the prediction markets. First, we give a comprehensive overview on the prediction markets, including their theoretical basis, classification and applications. Second, we present how to design prediction markets based on smart contracts. Then, the algorithm of contracts implementation is proposed. Finally, in order to verify the effectiveness of the algorithm, an intra-enterprise prediction market is built based on a private blockchain. The experimental results show that the market can make accurate prediction for a particular event. In addition, the autonomy, self-sufficiency, and decentralization characteristics of blockchain make the prediction markets more efficient and robust.

关键词： Smart contracts Blockchain Consumer electronics Companies Voting Classification algorithms

来源：评论

学校读者我要写书评

暂无评论

Convolutional Neural Network Features Comparison Between Back-Propagation and Extreme Learning Machine

Convolutional Neural Network Features Comparison Between Bac...

引用

Chinese control Conference

作者： Atmane Khellal Hongbin Ma Qing Fei School of Automation Beijing Institute of Technology Beijing China State Key Laboratory of Intelligent Control and Decision of Complex Systems Beijing Institute of Technology Beijing China

Recently deep learning based architectures have been widely deployed in many problems of artificial intelligence. Among deep learning models, Convolutional Neural Networks (CNN) have been reported in numerous successful applications such as object recognition, and natural language processing. The convolutional neural networks are trained by back-propagating the classification error using the Back-Propagation (BP) algorithm, which requires a large amount of data and slows the training process. To overcome these difficulties, a new fast and accurate approach based on Extreme Learning Machine (ELM) to train any convolutional neural network has been proposed. The developed framework (ELM-CNN) is based on the concept of auto-encoding to learn the convolutional filters with biases, by reconstructing the normalized input and the intercept term. In this paper, systematic comparison with traditional back-propagation based training method (BP-CNN) has been made with respect to two aspects qualitative and quantitative. The experimental results on the popular MNIST dataset show that the ELM-CNN algorithm achieves competitive results in terms of generalization performance and up to 16 times faster than the back-propagation based training of CNN.

关键词： Training Convolutional neural networks Feature extraction Computer architecture Convolution Machine learning Encoding

来源：评论

学校读者我要写书评

暂无评论

Online Context-based Person Re-identification and Biometric-based Action Recognition for Service Robots 29

Online Context-based Person Re-identification and Biometric-...

引用

第29届中国控制与决策会议

作者： Ning An Shi-Ying Sun Xiao-Guang Zhao Zeng-Guang Hou State Key Laboratory of Management and Control for Complex Systems Institute of AutomationChinese Academy of Sciences University of Chinese Academy of Sciences

ISBN: (纸本)9781509046584

In this paper, we address the problem of person re-identification and action recognition for service robots,which undergoes lack of training dataset for model learning, reduction of feature set discriminative power in changing scenarios, and high complexity of the algorithm computation. An online context-based person re-identification algorithm is proposed, which learns the person model online without pre-collect dataset and adjusts the weight of features according to the context information. An online biometric-based action recognition algorithm is proposed, actions are recognized by simply matching the skeleton vectors extracted from five linkage mechanisms of human body. The proposed algorithms are evaluated on a service robot system, extensive experimental results show that they performs efficiently and effectively in various real-life scenarios.

关键词： person re-identification action recognition online processing human-robot interaction service robot

来源：评论

学校读者我要写书评

暂无评论

Feature Fusion Based Insulator Detection for Aerial Inspection 36

Feature Fusion Based Insulator Detection for Aerial Inspecti...

引用

第36届中国控制会议

作者： YAN Tiantian YANG Guodong YU Junzhi University of Chinese Academy of Sciences State Key Laboratory of Management and Control for Complex Systems Institute of Automation ofChinese Academy of Sciences

ISBN: (纸本)9781538629185

This paper presents a detection method of insulator stings for aerial inspection based on *** local sub-images of insulator strings are firstly collected from aerial videos and tagged to establish a training *** fusion feature is then composed by the histogram of oriented gradients（HOG） feature and local binary pattern（LBP） feature after the principal component analysis（PCA） dimension reduction separately.A training model is developed by SVM algorithm with the fusion *** the detection phase,threshold segmentation and morphological operation are adopted to preprocess the *** sliding window method is then used to search the candidate region and the non-maximum suppression（NMS） method is adopted to fuse the candidate ***,the position of the insulator strings can be calculated by linear *** the efficiency and the effectiveness of the proposed method are verified through experiments on locating the multi-angle insulator strings under complex backgrounds.

关键词： Insulator location feature-fusion power line inspection HOG LBP

来源：评论

学校读者我要写书评

暂无评论

Data-Driven Adaptive Dynamic Programming for Two-Player Nonzero-Sum Game 29

Data-Driven Adaptive Dynamic Programming for Two-Player Nonz...

引用

第29届中国控制与决策会议

作者： Qichao Zhang Dongbin Zhao Yafei Zhou The state Key Laboratory of Management and Control for Complex Systems Institute of AutomationChinese Academy of Sciences University of Chinese Academy of Sciences

ISBN: (纸本)9781509046584

In this paper, we propose a data-driven adaptive dynamic programming approach to solve the Hamilton-Jacobi(HJ) equations for the two-player nonzero-sum(NZS) game with completely unknown dynamics. First, the model-based policy iteration(PI) algorithm is given, where the knowledge of system dynamics is required. To relax this requirement,a data-driven adaptive dynamic programming(ADP) is proposed in this paper to solve the unknown nonlinear NZS game with only online data. Neural network approximators are constructed to approach the solution of the HJ equations. The online data is collected under the two initial admissible control policies. Then, the NN weights are updated based on the least-squares method using the collected online data repeatedly, which is a kind of the off-policy learning ***, a simulation example is provided to demonstrate the effectiveness of the proposed control scheme.

关键词： Adaptive dynamic programming data-driven neural network unknown dynamics off-policy

来源：评论

学校读者我要写书评

暂无评论

Move Like Humans:End-to-end Gaussian Process Regression Based Target Tracking control for Mobile Robots 36

Move Like Humans:End-to-end Gaussian Process Regression Base...

引用

第36届中国控制会议

ISBN: (纸本)9781538629185

In this paper,we address the problem of target tracking control for mobile robots with limited sensing *** end-to-end Gaussian process regression learning control method is proposed to transfer the human control experiences to the *** end-to-end learning architecture directly learns the control mapping from the original sensing input space to the final control output space in an human-like *** non-parametric Gaussian process regression accurately transfers the complex human control experiences to the control *** addition,realistic training data set are collected from human operators for the end-to-end control *** performance of the proposed target tracking control method is extensively evaluated on various real-world scenarios,experimental results have demonstrated the robustness,accuracy,and effectiveness of the proposed method.

关键词： Target Tracking control End-to-end Learning Human-like control Non-parametric Bayesian Model Gaussian Process Regression

来源：评论

学校读者我要写书评

暂无评论

Research on Astronomical Trajectory Planning Algorithm for FAST 29

Research on Astronomical Trajectory Planning Algorithm for F...

引用

第29届中国控制与决策会议

作者： Sai Deng Fengshui Jing Guodong Yang Zize Liang State Key Laboratory of Management and Control for Complex Systems Institute of AutomationChinese Academy of Sciences University of Chinese Academy of Sciences

ISBN: (纸本)9781509046584

This paper addresses astronomical trajectory planning algorithms for tracking and basket-weaving for Five-hundred-meter Aperture Spherical Radio Telescope(FAST).In order to achieve stable operation at feed receiver of FAST and to ease difficulties of trajectory planning in 3-D space,the observed trajectories were planned in celestial coordinate ***,the characteristics of tracking and basket-weaving were analyzed,and the involved parameters for tracking and basket-weaving were ***,the constraint conditions for planning parameters corresponding to the practical motion were obtained;besides trajectory planning algorithms based on double S velocity profile for tracking and basket-weaving were ***,the numerical results indicated that the planned trajectories are guaranteed to be continuous in position,velocity and acceleration,as well as respect the given motion constraint *** the proposed algorithm cost less planning time than the algorithm plans in 3-D space.

关键词： trajectory planning FAST telescope double S

来源：评论

学校读者我要写书评

暂无评论

Cascade Motion Planning for Vibration Suppression of the Focus Cabin in FAST 36

Cascade Motion Planning for Vibration Suppression of the Foc...

引用

第36届中国控制会议

作者： Zheng Rongzhang Jing Fengshui Yang Guodong University of Chinese Academy of Sciences The State Key Laboratory of Management and Control for Complex Systems Institute of AutomationChinese Academy of Sciences

ISBN: (纸本)9781538629185

The Five-hundred-meter Aperture Spherical Radio Telescope（FAST） requires high accuracy of positioning and attitude *** to its flexible structure,the cable-cabin system can be excited to vibrate,which affects the *** alleviate the vibration during the slewing task,taking advantage of s-curve planning and input shaping,a cascade motion planning method is *** the main vibration mode,a simplified cable-cabin model is built for simulation according to the parameters of *** experiments are conducted to prove that the cascade method is effective in vibration suppression and robust to the change of natural frequency.

关键词： FAST telescope vibration suppression motion planning s-curve input shaping

来源：评论

学校读者我要写书评

暂无评论

Rethinking temporal object detection from robotic perspectives

arXiv

引用

arXiv 2019年

作者： Chen, Xingyu Wu, Zhengxing Yu, Junzhi Wen, Li State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences Beijing100190 China University of Chinese Academy of Sciences Beijing100049 China Beijing Innovation Center Engineering Science and Advanced Technology Peking University Beijing100871 China School of Mechanical Engineering and Automation Beihang University Beijing100191 China

Video object detection (VID) has been vigorously studied for years but almost all literature adopts a static accuracy-based evaluation, i.e., average precision (AP). From a robotic perspective, the importance of recall continuity and localization stability is equal to that of accuracy, but the AP is insufficient to reflect detectors' performance across time. In this paper, non-reference assessments are proposed for continuity and stability based on object tracklets. These temporal evaluations can serve as supplements to static AP. Further, we develop an online tracklet refinement for improving detectors' temporal performance through short tracklet suppression, fragment filling, and temporal location fusion. In addition, we propose a small-overlap suppression to extend VID methods to single object tracking (SOT) task so that a flexible SOT-by-detection framework is then formed. Extensive experiments are conducted on ImageNet VID dataset and real-world robotic tasks, where the superiority of our proposed approaches are validated and verified. Codes will be publicly available. Copyright © 2019, The Authors. All rights reserved.

关键词： Object recognition

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：