检索结果-南通市图书馆

Skill enhancement learning with knowledge distillation

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第8期67卷 206-220页

作者： Naijun LIU fuchun sun Bin FANG Huaping LIU Department of Computer Science and Technology Tsinghua University

Skill learning through reinforcement learning has significantly progressed in recent years. However, it often struggles to efficiently find optimal or near-optimal policies due to the inherent trial-and-error exploration in reinforcement learning. Although algorithms have been proposed to enhance skill learning efficacy, there is still much room for improvement in terms of skill learning performance and training stability. In this paper, we propose an algorithm called skill enhancement learning with knowledge distillation(SELKD), which integrates multiple actors and multiple critics for skill learning. SELKD employs knowledge distillation to establish a mutual learning mechanism among actors. To mitigate critic overestimation bias,we introduce a novel target value calculation method. We also perform theoretical analysis to ensure the convergence of SELKD. Finally, experiments are conducted on several continuous control tasks, illustrating the effectiveness of the proposed algorithm.

关键词： skill learning enhancement learning reinforcement learning knowledge distillation

同方期刊数据库

Data-driven electrical resistance tomography for robotic large-area tactile sensing

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第09期67卷 336-337页

作者： Wendong ZHENG Huaping LIU Xiaofeng LIU fuchun sun School of Electrical Engineering and Automation Tianjin University of Technology Department of Computer Science and Technology Tsinghua University Key Laboratory of Maritime Intelligent Cyberspace Technology (Hohai University) Ministry of Education

Tactile sensing plays a crucial role in enabling robots to safely interact with objects in dynamic environments [1].Given that potential physical contact can occur at any location during robot interaction, there is a ... 详细信息

关键词： Robot applications

同方期刊数据库

Small sample learning with high order contractive auto-encoders and application in SAR images

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2018年第9期61卷 306-318页

作者： Qianwen YANG fuchun sun Department of Computer Science and Technology Tsinghua University State Key Lab of Intelligent Technology and Systems Tsinghua University Tsinghua National Laboratory for Information Science and Technology Tsinghua University

Dear editor,Recently auto-encoders(AEs)are used as intermediate layers or unsupervised learning stages in deep learning networks[1].However,unlike other deep learning algorithms,which can extract higher-

关键词： SAR Small sample learning with high order contractive auto-encoders and application in SAR images MSTAR RBM

基于亚像素精度的汽车冲压件尺寸测量系统设计

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

组合机床与自动化加工技术 2024年第3期 44-48,53页

作者：吴昊荣李晓晓孙付春成都大学电子信息与电气工程学院成都610106 成都大学机械工程学院成都610106 四川三思联创传感技术有限公司成都610000

针对汽车冲压件尺寸人工测量方式存在的检测效率低和测量精度难以保证等问题,提出了一种基于亚像素精度的尺寸测量方法。先对汽车冲压件进行图像采集和预处理,接着利用Canny算子和改进Zernike矩进行冲压件图像亚像素边缘提取,再通过Houg... 详细信息

针对汽车冲压件尺寸人工测量方式存在的检测效率低和测量精度难以保证等问题,提出了一种基于亚像素精度的尺寸测量方法。先对汽车冲压件进行图像采集和预处理,接着利用Canny算子和改进Zernike矩进行冲压件图像亚像素边缘提取,再通过Hough变换对提取的圆轮廓进行拟合,最终获得了冲压件外圆直径和3个内孔直径的精确尺寸,4个圆轮廓的测量精度均高于0.025 mm,且多次测量结果的标准差小于0.009 6 mm。基于亚像素精度的尺寸测量系统,完成了测量平台的搭建和软件系统的开发,实现了汽车冲压件尺寸的非接触式亚像素级精度测量,能够满足工业应用的快速连续性测量要求。

关键词：汽车冲压件亚像素改进Zernike矩尺寸测量

维普期刊数据库

Routing for predictable Multi-Layered Satellite Networks

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2013年第11期56卷 18-35页

作者： LIU HeYu sun fuchun State Key Laboratory of Intelligence Technology and Systems Department of Computer Science and Technology Tsinghua University

Multi-Layered Satellite Network(MLSN),especially LEO/MEO satellite network,consisting of low and medium earth orbit satellites,is capable of providing higher coverage and better service than single-layered *** performance in deployment,however,is mainly encumbered by obsolete routing protocols and *** paper takes the predictability of satellite movements into consideration,based on which a novel routing protocol Predictable Satellite Network Routing Protocol(PSNRP)—is *** this protocol,every topology change due to satellite movement is either described as a predictable or an unpredictable *** the basis of PSNRP,we import a heuristic routing algorithm for satisfying Quality-of-Service(QoS)requirements from network users,which is convinced as a Non-deterministic Polynomial Complete *** QoS requirements in this paper mainly refer to end-to-end delay,bandwidth and package loss *** addition to obtaining better routing performance,PSNRP successfully allocates calculation tasks evenly among all nodes,separates user data from protocol control data and achieves stronger robustness against satellite failures and link *** superiority of our routing scheme is proved by simulation results.

关键词： Multi-Layered Satellite Nework LEO/MEO routing predictable heuristic Quality-of-Service robustness

A statistical learning based image denoising approach

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2015年第5期9卷 713-719页

作者： Ke TU Hongbo LI fuchun sun Department of Computer Science and Technology State Key Laboratory of Intelligent Technology and Systems Tsinghua University Beijing 100084 China

The image denoising is a very basic but important issue in the field of image procession. Most of the existing methods addressing this issue only show desirable performance when the image complies with their underlying assumptions. Especially, when there is more than one kind of noises, most of the existing methods may fail to dispose the corresponding image. To address this problem, we propose a two-step image denoising method motivated by the statistical learning theory. Under the proposed framework, the type and variance of noise are estimated with support vector machine （SVM） first, and then this information is employed in the proposed denoising algorithm to further improve its denoising performance. Finally, comparative study is constructed to demonstrate the advantages and effectiveness of the proposed method.

关键词： SVM image denosing multiple noises

维普期刊数据库

An effective method for grasp planning on objects with complex geometry combining human experience and analytical approach

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2016年第11期59卷 170-181页

作者： Chunfang LIU fuchun sun Xiaojuan BAN Department of Computer Science and Technology Tsinghua University Department of Computer Science and Technology University of Science and Technology Beijing

In this paper, an effective method for identifying the graspable components of objects with complex geometry is proposed for grasp planning based on human experience. Instead of focusing on individual objects,our method identifies graspable components on the category level under the assumption that geometrically alike objects share similar graspable components. Firstly, employing a modified SHOT descriptor, a fast KNN-based method is developed for object categorization. Then, the graspable components are identified by adopting a learning framework based on human experience. Afterwards, a fast analytical grasp planning method is proposed which comprises of contact points exaction and hand kinematics calculation. Finally, a regression model based on the extreme learning method(ELM) is built which inputs the desired contact points and the wrist orientation and outputs the wrist position. This approach is time-saving comparing with the optimization method. The simulations and experiments demonstrate the effectiveness of the proposed approach by realizing grasps on the graspable components of human choice for objects with complex geometry.

关键词： grasp planning human experience analytical method kinematics learning ELM

Virtual Strategy QoS routing in satellite networks

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2016年第9期59卷 79-90页

作者： Heyu LIU fuchun sun Shaoqing WANG State Key Laboratory of Intelligence Technology and Systems Department of Computer Science and TechnologyTsinghua University

Satellite network users want to unify different satellites to enhance the quality-of-service(Qo S)stability of the satellite networks and select feasible paths through different networks to feed different applications. Unfortunately, the state of affairs is that different applications need to use several different application programming interfaces and to design different protocols on how and when to use a specific network. This is troublesome and error-prone as the application programming interfaces vary a lot. In this paper, we design a Virtual Strategy in satellite network based on which a Qo S routing service scheme is then proposed. We analyze why applications should use and benefit from Virtual Strategy. This Virtual Strategy is a middleware solution that enables seamless usage of services from different satellite network parts. And then, the supporting Qo S routing solution enables the committed Qo S services over Virtual Strategy. Finally, we provide a comparison between the previous satellite networks and our work. The simulation results show that our Virtual Strategy Qo S routing scheme demonstrates dominated performances under complex architecture.

关键词： satellite network QoS services Virtual Strategy middleware routing scheme

Experiment on impedance adaptation of under-actuated gripper using tactile array under unknown environment

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2018年第12期61卷 127-140页

作者： Jing CUI Mi LAI Zhongyi CHU fuchun sun School of Mechanical Engineering and Applied Electronics Beijing University of Technology Department of Computer Science and Technology Tsinghua University School of Instrument Science and Opto-electronics Beihang University

The experiment on impedance adaptation to achieve stable grasp for an under-actuated gripper grasping different unknown objects with tactile array is conducted. Under-actuated gripper has a wildly application in the field of space robot and industrial robot because of its better shape-adaptation. However it is difficult to achieve stable grasp owning to the uncertain properties of environment. A control strategy of adaptive matching the impedance parameters is proposed to achieve stable grasp. Firstly, the unknown objects are described as linear systems with unknown dynamics, and the parameters of the object are identified with the recursive least-squares(RLS) method through tactile sensor array. Then a desired impedance model is obtained by defining a cost function that includes the contact force, velocity and displacement errors, and the critical impedance parameters are found to minimize it. Finally, an experiment is presented and shows that the proposed impedance model can guarantee the stable grasp for various unknown objects.

关键词： impedance adaptation under-actuated gripper tactile array stable grasp unknown environment