检索结果-内蒙古大学图书馆

arXiv 2020年

作者： Wang, Junjie Zhang, Qichao Zhao, Dongbin Zhao, Mengchen Hao, Jianye State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences Beijing China Huawei Noah’s Ark Lab Beijing China

Existing model-based value expansion methods typically leverage a world model for value estimation with a fixed rollout horizon to assist policy learning. However, the fixed rollout with an inaccurate model has a potential to harm the learning process. In this paper, we investigate the idea of using the model knowledge for value expansion adaptively. We propose a novel method called Dynamic-horizon Model-based Value Expansion (DMVE) to adjust the world model usage with different rollout horizons. Inspired by reconstruction-based techniques that can be applied for visual data novelty detection, we utilize a world model with a reconstruction module for image feature extraction, in order to acquire more precise value estimation. The raw and the reconstructed images are both used to determine the appropriate horizon for adaptive value expansion. On several benchmark visual control tasks, experimental results show that DMVE outperforms all baselines in sample efficiency and final performance, indicating that DMVE can achieve more effective and accurate value estimation than state-of-the-art model-based methods. Copyright © 2020, The Authors. All rights reserved.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Parallel Hospital: ACP-Based Hospital Smart Operating System

Parallel Hospital: ACP-Based Hospital Smart Operating System

引用

Digital Twins and Parallel Intelligence (DTPI), IEEE International Conference on

作者： Yongjun Wang Fei-Yue Wang Xiao Wang Zhongjun Guan Liwei Ouyang Jiangong Wang Lan Yan Wenbo Zheng Wenwen Zhang Beijing Tiantan Hospital Capital Medical University Beijing Beijing China The State Key Laboratory for Management and Control of Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China Qingdao Academy of Intelligent Industries Qingdao China School of Artificial Intelligence University of Chinese Academy of Sciences Beijing China School of Software Engineering Xi'an Jiaotong University Xi'an China

This article aims to propose a parallel hospital based on parallel medicine. The parallel hospital takes the parallel idea of virtual and real interaction as its core concept and uses parallel idea's artificial scenarios, computational experiments, and parallel execution (ACP) theory as the guiding theory to construct a new generation of intelligent hospital management system with virtual and real interaction. We conducted a case study of the parallel hospital practice case of Tiantan Hospital-Tiantan Smart Brain, so as to confirm the feasibility and scientificity of the proposed parallel hospital.

关键词： Training Hospitals Operating systems Digital twin Conferences

来源：评论

学校读者我要写书评

暂无评论

Programming by Visual Demonstration for Pick-and-Place Tasks using Robot Skills

Programming by Visual Demonstration for Pick-and-Place Tasks...

引用

IEEE International Conference on Robotics and Biomimetics

作者： Peng Hao Tao Lu Yinghao Cai Shuo Wang The State Key Laboratory for Management and Control of Complex Systems Chinese Academy of Sciences Beijing China

In this paper, we present a vision-based robot programming system for pick-and-place tasks that can generate programs from human demonstrations. The system consists of a detection network and a program generation module. The detection network leverages convolutional pose machines to detect the key-points of the objects. The network is trained in a simulation environment in which the train set is collected and auto-labeled. To bridge the gap between reality and simulation, we propose a design method of transform function for mapping a real image to synthesized style. Compared with the unmapped results, the Mean Absolute Error (MAE) of the model completely trained with synthesized images is reduced by 23% and the False Negative Rate FNR (FNR) of the model fine-tuned by the real images is reduced by 42.5% after mapping. The program generation module provides a human-readable program based on the detection results to reproduce a real-world demonstration, in which a longshort memory (LSM) is designed to integrate current and historical information. The system is tested in the real world with a UR5 robot on the task of stacking colored cubes in different orders.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Acupoint Detection Based on Deep Convolutional Neural Network

Acupoint Detection Based on Deep Convolutional Neural Networ...

引用

第三十九届中国控制会议

作者： Lingyao Sun Shiying Sun Yuanbo Fu Xiaoguang Zhao State Key Laboratory of Management and Control for Complex Systems Institute of AutomationChinese Academy of Sciences University of Chinese Academy of Sciences school of Artificial Intelligence The Department of Acupuncture and Moxibustion Beijing Hospital of Traditional Chinese MedicineCapital Medical UniversityBeijing Key Laboratory of Acupuncture Neuromodulation

As an important component of Traditional Chinese Medicine(TCM),science of acupoint therapy has achieved significant results in clinical practice,but recognizing and positioning acupoints is heavily depends on the skills of *** recent years,researchers have proposed a few methods of automatic acupoints detection and positioning,but most of the methods are still based on manual designed *** this paper,we propose an acupoints detection method based on deep convolutional neural network,and an evaluation method is proposed for acupoint ***'s more,we build an acupoint detection *** are performed and a promising result is achieved.

关键词： convolutional neural network deep learning acupoint detection convolutional pose machine

来源：评论

学校读者我要写书评

暂无评论

Line-of-Sight Strategy-Based Path-Following System for a Multi-Joint Robotic Fish

Line-of-Sight Strategy-Based Path-Following System for a Mul...

引用

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

作者： Shijie Dai Chao Zhou Zhengxing Wu Min Tan Junzhi Yu School of Artificial Intelligence University of Chinese Academy of Sciences Beijing China State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China Naval Research Academy Beijing China State Key Laboratory for Turbulence and Complex System BIC-ESAT College of Engineering Peking University Beijing China

ISBN: (数字)9781728172934

ISBN: (纸本)9781728172941

This paper proposes a real-time path-following control system for a multi-joint robotic fish. The mechanical structure and dynamic model of the robotic fish for pathfollowing are first described. Then, the framework of the path-following control algorithm is established based on the built dynamic model, including a modified line-of-sight (LOS) guidance law, an active disturbance rejection control (ADRC)based heading controller and a proportional-integral-derivative (PID)-based speed controller. Specially, the modified LOS strategy is designed to select the tracking points and also provide the desired heading angle. Afterwards, to overcome systematic uncertainties and environmental disturbances, the ADRC method is adopted to design the heading controller. Meanwhile, the PID controller is also developed to maintain an appropriate swimming speed. Finally, simulations in both linear-and circular-path following are presented to validate the effectiveness of the proposed method.

关键词： Robots Robot kinematics Task analysis Real-time systems Oscillators Mathematical model Force

来源：评论

学校读者我要写书评

暂无评论

A Hybrid Deep Learning Approach with GCN and LSTM for Traffic Flow Prediction*

A Hybrid Deep Learning Approach with GCN and LSTM for Traffi...

引用

International Conference on Intelligent Transportation

作者： Zhishuai Li Gang Xiong Yuanyuan Chen Yisheng Lv Bin Hu Fenghua Zhu Fei-Yue Wang State Key Laboratory for Management and Control of Complex Systems Chinese Academy of Sciences Beijing China

ISBN: (纸本)9781538670255

Traffic flow prediction is an important functional component of Intelligent Transportation systems (ITS). In this paper, we propose a hybrid deep learning approach, called graph and attention-based long short-term memory network (GLA), to efficiently capture the spatial-temporal features in traffic flow. Firstly, we apply graph convolutional network (GCN) to mine the spatial relationships of traffic flow over multiple observation stations, in which the adjacent matrix is determined by a data-driven approach. Then, we feed the output of the GCN model to the long short-term memory (LSTM) model which extracts temporal features embedded in traffic flow. Further, we implement a soft attention mechanism on the extracted spatial-temporal traffic features to make final prediction. We test the proposed method over the PeMS data sets. Experimental results show that the proposed model performs better than the competing methods.

关键词： Feature extraction Machine learning control systems complex systems Automation Predictive models Data models

来源：评论

学校读者我要写书评

暂无评论

ContourRend: A Segmentation Method for Improving Contours by Rendering

arXiv

引用

arXiv 2020年

作者： Chen, Junwen Lu, Yi Chen, Yaran Zhao, Dongbin Pang, Zhonghua 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 Beijing101408 China Key Laboratory of Fieldbus Technology and Automation of Beijing North China University of Technology Beijing100144 China

A good object segmentation should contain clear contours and complete regions. However, mask-based segmentation can not handle contour features well on a coarse prediction grid, thus causing problems of blurry edges. While contour-based segmentation provides contours directly, but misses contours' details. In order to obtain fine contours, we propose a segmentation method named ContourRend which adopts a contour renderer to refine segmentation contours. And we implement our method on a segmentation model based on graph convolutional network (GCN). For the single object segmentation task on cityscapes dataset, the GCN-based segmentation contour is used to generate a contour of a single object, then our contour renderer focuses on the pixels around the contour and predicts the category at high resolution. By rendering the contour result, our method reaches 72.41% mean intersection over union (IoU) and surpasses baseline Polygon-GCN by 1.22%. Copyright © 2020, The Authors. All rights reserved.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Fast and Robust RGB-D Multiple Human Tracking Based on Part Model for Mobile Robots

Fast and Robust RGB-D Multiple Human Tracking Based on Part ...

引用

第三十八届中国控制会议

作者： Shiying Sun Xiaoguang Zhao Min Tan State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences

In this paper, we address the problem of multiple human tracking for mobile robots and proposed a fast and robust RGB-D multiple human tracking approach based on part model. Firstly, a simplified deformable part model method is used for fast human detection using RGB-D information. Then the partial occlusion is detected with depth information. Finally, part model and partial occlusion information are combined into data association process and an on-line appearance classification method with dynamic occlusion handling is proposed for robust tracking. The performance of the proposed method is evaluated on the public dataset, and experiment results have demonstrated the effectiveness of the proposed method.

关键词： Multiple Human Tracking Deformable Part Model Human Detection Occlusion Handling On-line Classification

来源：评论

学校读者我要写书评

暂无评论

Learning Category-level Implicit 3D Rotation Representations for 6D Pose Estimation from RGB Images

Learning Category-level Implicit 3D Rotation Representations...

引用

IEEE International Conference on Robotics and Biomimetics

作者： Xiaocan Li Yinghao Cai Shuo Wang Tao Lu State Key Laboratory of Management and Control for Complex Systems Chinese Academy of Sciences Beijing China

We exploit the embedding ability of a de-noising autoencoder for an implicit 3D rotation representation learning at the category level. Contrast to the exact 3D reconstruction model of each instance-level physical object, we leverage the inexact CAD/Reconstruction models of an object as the representative model for some category. Under our assumptions that objects within the same category share resemble geometry, we train a de-noising autoencoder on synthetic 3D views of category-level objects to extract the homogenous features at the bottleneck layer. The latent representation is agnostic not only to heterogeneous textures, colors, and illuminations, but also ambiguous pose caused by object symmetry. To extend the instance-level 3D translation estimation to the category level, we considered the 3D diagonal length ratio between the source and target object. We achieved a frame rate of 17Hz.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Feature Aggregation Attention Network for Single Image Dehazing

Feature Aggregation Attention Network for Single Image Dehaz...

引用

IEEE International Conference on Image Processing

作者： Lan Yan Wenbo Zheng Chao Gou Fei-Yue Wang The State Key Laboratory for Management and Control of Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China School of Software Engineering Xi’an Jiaotong University Xi’an China School of Intelligent Systems Engineering Sun Yat-sen University Guangzhou China

ISBN: (数字)9781728163956

ISBN: (纸本)9781728163963

Due to its ill-posed nature, single image dehazing is a challenging problem. In this paper, we propose an end-to-end feature aggregation attention network (FAAN) for single image dehazing. It incorporates the idea of attention mechanism and residual learning and can adaptively aggregate different level features. In particular, in the proposed FANN, we design a novel block structure consisting of feature attention module, smoothed dilated convolution and local residual learning. The local residual learning allows the less useful information to be bypassed through multiple skip connections. The feature attention module is designed to assign more weight to important features. The smoothed dilated convolution is adopted to enlarge the receptive field without the negative influence of gridding artifacts. The experiments on the RESIDE dataset show that the proposed approach acquires state-of-the-art performance in both qualitative and quantitative measures.

关键词： Convolution Atmospheric modeling Kernel Image color analysis Task analysis Training

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：