检索结果-内蒙古大学图书馆

American Control Conference

作者： Samet Guler Mohamed Abdelkader Jeff S. Shamma Computer Electrical and Mathematical Science and Engineering Division (CEMSE) KAUST Robotics Intelligent Systems and Control (RISC) Lab Thuwal 23955-6900 Saudi Arabia

ISBN: (纸本)9781538679012;9781538679265

Swarm applications use motion capture system or GPS sensors as localization systems. However, motion capture systems provide local solutions, and GPS sensors are not reliable in occluded environments. For reliable and versatile operation in a swarm, robots must sense each other and interact locally. Motivated by this requirement, we propose an onboard localization framework for multi-robot systems. Our framework consists of an anchor robot with three ultrawideband (UWB) sensors and a tag robot with a single UWB sensor. The anchor robot utilizes the three UWB sensors as a localization infrastructure and estimates the tag robot's location by using its on-board sensing and computational capabilities solely, without explicit inter-robot communication. We utilize a dual Monte-Carlo localization approach to capture the agile maneuvers of the tag robot with acceptable precision. We validate the effectiveness of our algorithm with simulations as well as indoor and outdoor experiments on a two-drone setup. The proposed framework with the dual MCL algorithm yields accurate estimates for various speed profiles of the tag robot, outperforms the standard particle filter and extended Kalman filter, and suffice for a relative position maintenance application.

关键词： Ultra wideband technologies Robot motion capture Sensors Particle filters multi-robot systems Extended Kalman filters Frameworks

来源：评论

学校读者我要写书评

暂无评论

Mesoscopic magnetic systems: From fundamental properties to devices

arXiv

引用

arXiv 2021年

作者： Heyderman, Laura J. Grollier, Julie Marrows, Christopher H. Vavassori, Paolo Grundler, Dirk Makarov, Denys Pané, Salvador Laboratory for Mesoscopic Systems Department of Materials ETH Zurich Zurich8093 Switzerland Laboratory for Multiscale Materials Experiments Paul Scherrer Institute Villigen PSI5232 Switzerland Unité Mixte de Physique CNRS Thales Université Paris-Saclay Palaiseau91767 France School of Physics & Astronomy University of Leeds LeedsLS2 9JT United Kingdom CIC nanoGUNE BRTA 20018 San Sebastian IKERBASQUE Basque Foundation for Science Bilbao48009 Spain Lausanne1015 Switzerland Helmholtz-Zentrum Dresden-Rossendorf e.V. Institute of Ion Beam Physics and Materials Research Dresden01328 Germany Multi-Scale Robotics Lab Institute of Robotics and Intelligent Systems ETH Zürich Tannenstrasse 3 ZürichCH 8092 Switzerland

来源：评论

学校读者我要写书评

暂无评论

Real-world Robot Reaching Skill Learning Based on Deep Reinforcement Learning

Real-world Robot Reaching Skill Learning Based on Deep Reinf...

引用

第32届中国控制与决策会议

作者： Naijun Liu Tao Lu Yinghao Cai Rui Wang Shuo Wang State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences University of Chinese Academy of Sciences Guangdong Provincial Key Lab of Robotics and Intelligent System Shenzhen Institutes of Advanced TechnologyChinese Academy of Sciences Center for Excellence in Brain Science and Intelligence Technology of the Chinese Academy of Sciences

Traditional programming method can achieve certain manipulation tasks with the assumption that robot environment is known and ***,with robots gradually applied in more domains,robots often encounter working scenes which are complicated,unpredictable,and *** overcome the limitation of traditional programming method,in this paper,we apply deep reinforcement learning(DRL) method to train robot agent to obtain skill *** policy trained with DRL on real-world robot is time-consuming and costly,we propose a novel and simple learning paradigm with the aim of training physical robot ***,our method train a virtual agent in an simulated environment to reach random target position from random initial ***,virtual agent trajectory sequence obtained with the trained policy,is transformed to real-world robot command with coordinate transformation to control robot performing reaching *** show that the proposed method can obtain self-adaptive reaching policy with low training cost,which is of great benefits for developing intelligent and robust robot manipulation skill system.

关键词： Robot reaching skill learning deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Obstacle avoidance with reinforcement learning and adaptive resonance theory

Obstacle avoidance with reinforcement learning and adaptive ...

引用

2019 IEEE International Conference on robotics and Biomimetics, ROBIO 2019

作者： Ye, Lingjian Zhou, Yimin Chinese Academy of Sciences Shenzhen Institutes of Advanced Technology Shenzhen518055 China Guangdong Provincial Key Lab of Robotics and Intelligent System Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences China CAS Key Laboratory of Human-Machine Intelligence-Synergy Systems Shenzhen Institutes of Advanced Technology Shenzhen China

ISBN: (纸本)9781728163215

The reinforcement learning (RL) of the autonomous mobile agent is one of the actual research topics. It permits mobile agents to interact constantly with their environment and to avoid obstacles. First, this paper presents an algorithm which integrates Deep Deterministic Policy Gradient (DDPG) algorithm and Fuzzy Adaptive Resonance Theory (ART) in order to improve generalization performance of RL. Then the curiosity is introduced to integrate with the first proposed algorithm to solve the problem of slow convergence caused by Fuzzy ART. Results of the simulation experiments demonstrate the effectiveness of the proposed algorithm. It shows that the algorithms perform well in both low-dimension and high-dimension state space. © 2019 IEEE.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

A Novel Multi-scale Fourier Descriptor Based on Plane Orthogonal Vectors for Fast Shape Retrieval

A Novel Multi-scale Fourier Descriptor Based on Plane Orthog...

引用

IEEE International Conference on robotics and Biomimetics

作者： Jiajun Ma Jiafeng Zhu Guoyuan Liang Xinyu Wu Guangdong Provincial Key Lab of Robotics and Intelligent System & CAS Key Laboratory of Human-Machine Intelligence-Synergy Systems Chinese Academy of Sciences Shenzhen China

Shape retrieval is an interesting and challenging issue in computer vision and robot vision. The recent works employing Fourier-based multi-scale descriptors achieve some success in shape retrieval, especially make real-time application possible by improving the efficiency. However, these methods are commonly adopting really complex shape descriptors to present the shape information which is not intuitive enough. In our work, we propose a new Fourier-based method with novel multi-scale descriptor which is a combination of contour feature and regional feature. This novel descriptor has two main characteristics: First, utilize six different invariant signatures which are based on plane orthogonal vectors; Second, leverage multiple scale structures to describe local and global information of shapes. Except the easiness in understanding, this descriptor approach also produces a more distinct topology of shapes than other approaches which means the property of stability on variations, linear or nonlinear. The method is tested on some popular databases that provides a promising retrieval rate(87.46% and 99.27%) with constant time complexity optimization. The results confirm that our method is more capable of representing the topology of shapes than most other methods based on Fourier descriptor.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A method of cliff detection in robot navigation based on multi-sensor

A method of cliff detection in robot navigation based on mul...

引用

IEEE International Conference on Real-time Computing and robotics (RCAR)

作者： Zhilong Su Can Wang Yuxiao Li Xinyu Wu Guangdong Provincial Key Lab of Robotics and Intelligent System Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences Shenzhen China CAS Key Laboratory of Human-Machine Intelligence Synergic Systems Shenzhen Institutes of Advanced Technolog Shenzhen China Shenzhen College of Advanced Technology University of Chinese Academy of Sciences Beijing China

ISBN: (数字)9781728172934

ISBN: (纸本)9781728172941

This paper proposed a method to detect cliff edge by using an RGB-D camera, and then to fuse the cliff information with LiDAR, finally to provide navigation information for robot autonomous exploration and autonomous movement. Compared with the prior methods which set sonar or infra-red sensor in the bottom of the robot, our method could detect cliff earlier by set the RGB-D camera forward. In cliff detection, we use region growing algorithm and the normal vector of plane to segment out ground plane in point cloud, then extract the boundary points of ground, and we proposed an Edge Point Filter (EPF) to get the edge points of cliff, then use the edge points to simulate LiDAR data and combine with real LiDAR data to mark the untraversable area. We verified the practicability of this method in simulation, and verified the accuracy with real sensors.

关键词： Laser radar Robots Three-dimensional displays Robot sensing systems Indexes Cameras Image edge detection

来源：评论

学校读者我要写书评

暂无评论

Trajectory-based off-policy deep reinforcement learning

arXiv

引用

arXiv 2019年

作者： Doerr, Andreas Volpp, Michael Toussaint, Marc Trimpe, Sebastian Daniel, Christian Bosch Center for Artificial Intelligence Renningen Germany Max Planck Institute for Intelligent Systems Stuttgart/Tubingen Germany Machine Learning and Robotics Lab University of Stuttgart Germany

Policy gradient methods are powerful reinforcement learning algorithms and have been demonstrated to solve many complex tasks. However, these methods are also data-inefficient, afflicted with high variance gradient estimates, and frequently get stuck in local optima. This work addresses these weaknesses by combining recent improvements in the reuse of off-policy data and exploration in parameter space with deterministic behavioral policies. The resulting objective is amenable to standard neural network optimization strategies like stochastic gradient descent or stochastic gradient Hamiltonian Monte Carlo. Incorporation of previous rollouts via importance sampling greatly improves data-efficiency, whilst stochastic optimization schemes facilitate the escape from local optima. We evaluate the proposed approach on a series of continuous control benchmark tasks. The results show that the proposed algorithm is able to successfully and reliably learn solutions using fewer system interactions than standard policy gradient methods. Copyright © 2019, The Authors. All rights reserved.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Low Light Enhancement by Unsupervised Network*

Low Light Enhancement by Unsupervised Network*

引用

IEEE International Conference on Real-time Computing and robotics (RCAR)

作者： Yangyang Qu Yongsheng Ou Rong Xiong Chinese Academy of Sciences Shenzhen Institutes of Advanced Technology Shenzhen China University of Chinese Academy of Sciences Guangdong Provincial Key Lab of Robotics and Intelligent System Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences Shenzhen China CAS Key Laboratory of Human-Machine Intelligence Synergic Systems Shenzhen Institutes of Advanced Technology Shenzhen China

ISBN: (数字)9781728172934

ISBN: (纸本)9781728172941

This paper focuses on unsupervised low-light enhancement methods for practical applications such as automated driving, surveillance cameras, and so on. This paper proposes an unsupervised method that is independent of paired samples and works well to solve the image darkness problem. Furthermore, to increase the brightness of the image and restore the information hidden in the dark, an auxiliary attention module is incorporated into our model. This proposed module can help generate images with good edges and make the image colors and details richer. The experimental results show that the proposed method works well for both image enhancement and noise control.

关键词： Image edge detection Image enhancement Lighting Generators Visual effects Histograms Brightness

来源：评论

学校读者我要写书评

暂无评论

A method for generating large-scale high definition color-point map

A method for generating large-scale high definition color-po...

引用

2019 IEEE International Conference on Real-Time Computing and robotics, RCAR 2019

作者： Li, Yuxiao Wang, Can Chen, Fan Su, Zhilong Wu, Xinyu Guangdong Provincial Key Lab of Robotics and Intelligent System Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences Shenzhen518055 China CAS Key Laboratory of Human-Machine Intelligence Synergic Systems Shenzhen Institutes of Advanced Technolog Shenzhen518055 China Shenzhen College of Advanced Technology University of Chinese Academy of Sciences Beijing100049 China

ISBN: (纸本)9781728137261

We propose a method to reconstruct a High-Definition Color-Pointed map from the large-scale urban road scene. In contrast to prior methods, we collect the urban road data under the severe traffic conditions and several kilometers long data sets. In our High-Definition Color-Pointed map, we can clearly see the lane surface and surrounding environment. Moreover, our maps are collected in a high speed of 40km/h, which greatly improved the practicality of the whole system. The High-Definition Color-Pointed map can be used in perception, localization and automatic navigation in driverless cars, drones and smart city management. Images, lidar and gnss best poses are collected as the raw data. We use interpolation method to process the raw data, and use our algorithm to translate the Lidar poses into the images collected time. To detect the dynamic obstacle, we use the fast neural framework YOLOv2. Our method, Dislocation Projection, can solve the spare points problems. Our algorithm was evaluated on wide roads and narrow streets. The experimental results exhibited the effectiveness of the proposed approach. © 2019 IEEE.

关键词： Optical radar

来源：评论

学校读者我要写书评

暂无评论

Low illumination enhancement for object detection in self-driving

Low illumination enhancement for object detection in self-dr...

引用

2019 IEEE International Conference on robotics and Biomimetics, ROBIO 2019

作者： Qu, Yangyang Ou, Yongsheng Xiong, Rong Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences Shenzhen518055 China University of Chinese Academy of Sciences Beijing100049 China Guangdong Provincial Key Lab of Robotics and Intelligent System Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences China CAS Key Laboratory of Human-MachineIntelligence-Synergy Systems ShenzhenInstitutes of Advanced Technology China

ISBN: (纸本)9781728163215

Object detection plays an important role in the field of self-driving. Illumination has a great impact on object detection, but most of the current methods do not solve the problem of object detection in poor light environment well. We propose a network that can optimize the image conversion which is based on the Cycle Generative Adversarial Networks (CycleGAN). We redesign the discriminator network of Cycle-GAN, add additional discriminators, optimize multiple parts of the network such as loss functions, and add object detection networks after converting the network. The Robot Car dataset of Oxford University is utilized to verify the effectiveness of the proposed method, and the results proved that our method can significantly improve the detection accuracy and increase the detected object number in low illumination environment. © 2019 IEEE.

关键词： Image enhancement

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：