侯新文 - 中国科学院自动化研究所 -

个人简介

侯新文男硕导中国科学院自动化研究所侯新文，2001年毕业于北京大学数学系，获理学博士学位，同年进入南开大学数学系攻读博士后，出站后任中国科学院自动化研究所模式识别国家重点实验室副研究员，现为中科院自动化所项目研究员、硕士研究生导师。本人长期从事计算机视觉和机器学习研究，在人脸识别领域提出了直接形象模型(DAM),在子空间学习领域提出局部化非负矩阵分解(LNMF)，在图像文字检测中提出使用条件虽机场过滤文本部件，在物体检测、运动跟踪、流形学习、AdaBoost、机器学习基础理论等方面发表过多篇文章，GoogleScholar2200多次。近年来工作重点转向强化学习、生成对抗学习、元学习、智能博弈等通用人工智能方向，以在游戏博弈中战胜人类选手为应用目标，以提出人工智能的统一数学理论为长期研究目标。招生专业 081104-模式识别与智能系统招生方向计算机视觉模式识别理论与方法强化学习教育背景 1998-09--2001-06北京大学数学系博士工作简历 2004-07~现在,中国科学院自动化研究所,副研究员 2003-12~2004-06,中科院自动化所,项目聘用 2001-07~2003-11,南开大学数学系,博士后科研项目（1）自底向上的人机对抗技术,主持,市地级,2017-01--2019-12 （2）人机对抗空间建模与态势认知,参与,国家级,2016-11--2019-10 （3）2018年人机对抗挑战赛,参与,国家级,2018-03--2020-03 （4）2018年人机对抗挑战赛-先知兵圣AI接口与框架开发,主持,国家级,2018-03--2020-03

研究领域

模式识别、计算机视觉、深度学习、强化学习、生成对抗学习、智能博弈、人工智能的数学理论

近期论文

查看导师最新文章（温馨提示：请注意重名现象，建议点开原文通过作者单位确认）

期刊论文 [1]申翔翔，侯新文，尹传环，深度强化学习中状态注意力机制的研究，智能系统学报，2019. [2]D.Wang,X.W.Hou,J.Xu,S.Yue,C.-L.Liu,Trafficsigndetectionusingacascademethodwithfastfeatureextractionandsaliencytest,IEEETrans.IntelligentTransportationSystems,18(12):3290-3302,2017. [3]Jiang-ningWang,Xiao-linChen,Xin-wenHou,Li-bingZhou,Chao-DongZhuaandLi-qiangJia,Construction,implementationandtestingofanimageidentificationsystemusingcomputervisionmethodsforfruitflieswitheconomicimportance(Diptera:Tephritidae),PestManagementScience,WileyOnlineLibrary,73(7):1511-1528,2016. [4]Y.-M.Zhang,K.Huang,X.Hou,C.-L.Liu,Learninglocalitypreservinggraphfromdata,IEEETrans.SMCPartB,44(11):2088-2098,2014. [5]GuoqiangZhong,KaizhuHuang,XinwenHou,S.Xiang,LocalTangentSpaceLaplacianEigenmaps,ComputationalIntelligence,2012 [6]Y.-F.Pan,X.W.Hou,C.-L.Liu,Ahybridapproachtodetectandlocalizetextsinnaturalsceneimages,IEEETrans.ImageProcessing,20(3):800-813,2011. [7]张蕾,陈小琳,侯新文,刘成林,樊利民,汪兴鉴,实蝇科果实蝇属昆虫数字图像自动识别系统的构建和测试,昆虫学报，54(2):184-196,2011. [8]X.-B.Jin,C.-L.Liu,X.W.Hou,Regularizedmargin-basedconditionallog-likelihoodlossforprototypelearning,PatternRecognition,43(7):2428-2438,2010. [9]陈小琳,侯新文,刘成林,刘晓秋,张知彬，昆虫图像自动鉴别技术，昆虫知识，45(2):317-322,2008. [10]W.Dong,X.Hou,J.Liu,Y.Fang,C.JinandQ.Zhu,3DvirtualreconstructionofthepleistocenecheetahskullfromtheTangshan,Nanjing,China,ProgressInNaturalScience，vol.17(1),pp.74-79,2007. [11]董为，侯新文，房迎三，刘金毅，朱奇志，南京汤山早更新世猎豹头骨CT扫描数据的三维重建，自然科学进展，vol.16(4),pp.1146-1152,2006. [12]StanZ.Li,X.G.Lv,X.W.Hou,X.H.PengandQ.S.Cheng,LearningMultiviewFaceSubspacesandFacialPoseEstimation UsingIndependentComponentAnalysis,IEEETrans.ImageProcessing,14(6):705-712,2005. [13]ShuichengYan,XinwenHou,StanZ.Li,HongjiangZhang,andQianshengCheng,FaceAlignmentUsingView-BasedDirectAppearanceModels,SpecialIssueonFacialImageProcessing,AnalysisandSynthesis,InternationalJournalofImagingSystemsandTechnology,Vol.1,p106-112,2003. [14]侯新文，程乾生，一种改进的变分Snake模型，数学的理论与实践，vol.31(2),pp.202-205,2001. 专著章节 [1]A.Lu,X.W.Hou,C.-L.Liu,X.Chen,Insectrecognitionusingsparsecodinganddecisionfusion,In:ComputerVisionandPatternRecognitioninEnvironmentalInformatics,JunZhou,XiaoBaiandTerryCaelli(Eds.),IGIGlobal,pp.124-145,2015. 会议论文 [1]XiangxiangShen,ChuanhuanYin,YekunChaiandXinwenHou,ExponentialMovingAveragedQ-networkforDDPG,TheSecondChinapatternrecognitionandcomputervisionConference(PRCV),2019. [2]ZhunanLiandXinwenHou,MixingUpdateQ-valueforDeepReinforcementLearning,InternationalJointConferenceonNeuralNetworks(IJCNN),2019. [3]XiangxiangShen,ChuanhuanYin,XinwenHou,Self-AttentionforDeepReinforcementLeraning,InternationalConferenceonMathematicsandArtificialIntelligence(ICMAI),2019. [4]Y.Wang,X.-Y.Zhang,Y.Zhang,X.W.HouandC.-LLiu,ExploitingCoarse-to-FineMechanismforFine-GrainedRecognition,Proc.Int.Conf.ImageProcessing(ICIP),Phoenix,USA,September25-28,2016,pp.649-653. [5]D.Wang,X.W.Hou,C.-LLiu,TrafficSignDetectionfromVideo:AFastApproachwithTracking,Proc.3rdACPR,KualaLumpur,Malaysia,2015. [6]D.Wang,S.Yue,J.Xu,X.W.Hou,C.-LLiu,Asaliency-basedcascademethodforfasttrafficsigndetection,IEEEIntelligentVehiclesSymposium,Seoul,Korea,2015,pp.180-185. [7]Y.Liu,X.W.Hou,C.-L.Liu,Acompactspatialfeaturerepresentationforimageclassification,Proc.2ndACPR,Okinawa,Japan,2013,pp.601-605. [8]X.-J.Jin,Q.-F.Wang,X.W.Hou,C.-L.Liu,Visualgesturecharacterstringrecognitionbyclassification-basedsegmentationwithstrokedeletion,Proc.2ndACPR,Okinawa,Japan,2013,pp.120-124. [9]Y.Liu,X.-Y.Zhang,K.Huang,X.W.Hou,C.-L.Liu,Multipleoutlookslearningwithsupportvectormachines,Proc.ICONIP2012,LNCSVol.7665,pp.116-124. [10]A.Lu,X.W.Hou,C.-L.Liu,X.Chen,Insectspeciesrecognitionusingdiscriminativelocalsoftcoding,Proc.21thICPR,Tsukuba,Japan,2012,pp.1221-1224. [11]Y.-F.Pan,C.-L.Liu,X.W.Hou,Fastscenetextlocalizationbylearning-basedfilteringandverification,Proc.Int.Conf.ImageProcessing(ICIP),HongKong,2010,pp.2269-2272. [12]A.Lu,X.W.Hou,X.Chen,C.-L.Liu,Insectspeciesrecognitionusingsparserepresentation,Proc.BMVC2000,Aberystwyth,UK,2010. [13]X.-B.Jin,X.W.Hou,C.-L.Liu,Multi-classAdaBoostwithhypothesismargin,Proc.20thICPR,Istanbul,Turkey,2010,pp.65-68. [14]H.Wang,X.W.Hou,C.-L.Liu,Boostingincrementalsemi-superviseddiscriminantanalysisfortracking,Proc.20thICPR,Istanbul,Turkey,2010,pp.2748-2751. [15]Y.-M.Zhang,Y.Zhang,D.-Y.Yeung,C.-L.Liu,X.W.Hou,Transductivelearningonadaptivegraphs,Proc.24thAAAIConferenceonArtificialIntelligence,2010,pp.661-666. [16]G.Zhong,W.-J.Li,D.-Y.Yeung,X.W.Hou,C.-L.Liu,Gaussianprocesslatentrandomfield,Proc.24thAAAIConferenceonArtificialIntelligence,2010,pp.679-684. [17]G.Zhong,X.W.Hou,C.-L.Liu,Relativedistance-basedLaplacianeigenmaps,Proc.2009ChineseConferenceonPatternRecognition(CCPR)andFirstCJKWorkshoponPatternRecognition(CJKPR),Nanjing,China,2009,pp.784-788. [18]H.Wang,X.W.Hou,C.-L.Liu,Objecttrackingbybidirectionallearningwithfeatureselection,ICIP2009,Cairo,Egypt,2009. [19]Y.-M.Zhang,X.W.Hou,S.Xiang,C.-L.Liu,Subspaceregularization:anewsemi-supervisedlearningmethod,ECMLPKDD2009,Bled,Slovenia,LNCSVol.5782,Springer,2009,pp.586-601. [20]Y.-F.Pan,X.W.Hou,C.-L.Liu,Textlocalizationinnaturalsceneimagesbasedonconditionalrandomfield,Proc.10thICDAR,Barcelona,Spain,2009,pp.6-10. [21]X.-B.Jin,C.-L.Liu,X.W.Hou,Prototypelearningbymargin-basedconditionallog-likelihoodloss,Proc.19thICPR,Tampa,USA,2008. [22]Y.-F.Pan,X.W.Hou,C.-L.Liu,Arobustsystemtodetectandlocalizetextsinnaturalsceneimages,Proc.8thIAPRInt.WorkshoponDAS,Nara,Japan,Sep.2008,pp.35-42. [23]X.-H.Liu,C.-L.Liu,X.W.Hou,Apooledsubspacemixturedensitymodelforpatternclassificationinhigh-dimensionalspaces,Proc.IJCNN2008,HongKong,pp.2467-2472. [24]X.-B.Jin,X.Hou,C.-L.Liu,Ahybridgenerative-discriminativelearningalgorithmforBayesiannetworkstructure,Proc.5thInt.Conf.onWaveletAnalysisandPatternRecognition,Beijing,China,2007,Vol.2,pp.618-623. [25]FeiWu,Yong-GeWangandXin-WenHou,LicensePlateCharacterRecognitionBasedonFramelet,Proc.5thInt.Conf.onWaveletAnalysisandPatternRecognition,Beijing,China,2007,Vol.2,pp.673-676. [26]X.Hou,C.-L.Liu,TieniuTan,Learningboostedasymmetricclassifiersforobjectdetection,Proc.CVPR2006,NewYork,Vol.1,pp.330-338. [27]Y.Wang,X.W.Hou,T.N.Tan,RecognizeMulti-peopleInteractiveActivitybyPCA-HMMs,the7thAsianConferenceonComputerVision(ACCV06),LectureNotesinComputerScience,2006,vol.3581,pp.160-169. [28]J.L.Cui,T.N.Tan,X.W.Hou,Y.H.Wang,Z.S.Wei,AnIrisDetectionMethodBasedonStructureInformation,AdvancesinBiometricPersonAuthentication:InternationalWorkshoponBiometricRecognitionSystems,LectureNotesinComputerScience,2005,vol.3781,pp.157-164 [29]StanZ.Li,XianhuanPeng,XinwenHou,HongjiangZhangandQianshengCheng,Multi-viewfaceposeestimationbasedonsupervisedISAlearning,ProceedingsofthFifthIEEEInternationalConferenceonAutomaticFaceandGestureRecognition,2002. [30]XinwenHou,StanZ.Li,HongjiangZhang,andQianshengCheng,DirectAppearanceModels,IEEEInternationalConferenceonComputerVisionandPatternRecognition,pp.828-833,Vol1,2001. [31]S.Z.Li,X.W.Hou,H.J.Zhang,Q.S.Cheng,LearningSpatiallyLocalized,Parts-BasedRepresentation,ProceedingsofIEEEInternationalConferenceonComputerVisionandPatternRecognition,2001,vol.1,pp.207-212. [32]S.Z.Li,J.Yan,X.W.Hou,Z.Y.Li,andH.J.Zhang,LearningLowDimensionalInvariantSignatureof3-DObjectunderVaryingViewandIlluminationfrom2-DAppearances,InProceedingsof8thIEEEInternationalConferenceonComputerVision,Vancouver,Vol.1,pp.635–640,2001. [33]WeidongSun,andXinwenHou,Multi-SensorInformationSynthesizedMethodforCloudClearing,Proceedingsof1997China-JapanSymposiumonAdvancedInformationTechnology,1997,pp.362-367.