Autonomous Knowledge Acquisition_ICRA13_0825_MS

Autonomous Knowledge Acquisition_ICRA13_0825_MS

ID:40707788

大小:594.32 KB

页数:7页

时间:2019-08-06

Autonomous Knowledge Acquisition_ICRA13_0825_MS_第1页
Autonomous Knowledge Acquisition_ICRA13_0825_MS_第2页
Autonomous Knowledge Acquisition_ICRA13_0825_MS_第3页
Autonomous Knowledge Acquisition_ICRA13_0825_MS_第4页
Autonomous Knowledge Acquisition_ICRA13_0825_MS_第5页
资源描述:

《Autonomous Knowledge Acquisition_ICRA13_0825_MS》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库

1、CONFIDENTIAL.Limitedcirculation.Forreviewonly.1AutonomousKnowledgeAcquisitionbasedonActiveLearning:ApplicationtoHumanoidRobotsinIndoorEnvironmentDominikMaximiliánRamík,ChristopheSabourin,andKuroshMadaniAbstract—Inthispaper,wepresentacognitivesystemallow-arela

2、tivelysmallnumberofsamples.Thisimportantcapacityingarobottodiscovertheworld,inwhichitevolves,andtoisobservedinchildrenlearningthelanguage(seee.g.[6]).learnautonomouslyhighlevelknowledgeaboutitfromlowlevelThisproblemhasbeenaddressedondifferentdegreesinsensorda

3、ta.Thelearningisaccomplishedbothbyobservationandbyanactiveinteractionwithahuman,inwhichtherobotpreviousworks.Forexample,in[7]acomputationalmodelistheactorofthelearning.Wefirstdescribethesysteminofword-meaningacquisitionbyinteractionispresented.Inageneralmanner

4、andweprovideexperimentalresultsona[8]authorspresentacomputationalmodelforacquisitionhumanoidrobotinareal-worldenvironmentusingevery-dayofalexicondescribingsimpleobjectsinapopulationofobjects.Weshow,thatourapproachallowsahumanoidrobothumanoidrobots.In[9],ahuma

5、noidrobotistaughttotolearnsuccessfullywithoutnegativeinputandfromsmallnumberofsamples.associatesimpleshapestohumanlexicon.Interactivelearn-ingisexploredalsoin[10],wherearobotisrequiredtolearntodistinguishtwoclassesofobjects.In[11],aI.INTRODUCTIONhumanoidrobot

6、istaughtthroughadialogwithuntraineduserwiththeaimtolearndifferentobjectsandtograspNrecentyears,therehasbeenasubstantialprogressinthemproperly.Intheworkof[12],alexicalacquisitionIroboticsystemsabletorobustlyrecognizeobjectsinrealmodelispresentedcombiningmoretr

7、aditionalapproachesworldusingalargedatabaseofpre-collectedknowledgewiththeconceptofcuriositytoalternatetheattentionofthe(see[1]foranotableexample).Therehasbeen,however,learningrobot.Amoreadvancedworkusingaweakformcomparativelylessadvanceinautonomousacquisitio

8、nofofinteractionwiththetutorhasbeenrecentlypresentedinsuchaknowledge,wherewefacetwoimportantchallenges.[13],whereanonlinealgorithmisproposedallowingarobotOne,comingfromtheworlditself,isth

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。