http://www.informatik.uni-trier.de/~ley/db/conf/cvpr/cvpr2001-1.html
CVPR 2001 (vol 1)
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990634	125	A High-Resolution Panoramic Camera	acquisition advanced agents ahuja aiaa analysis anderson angle applied approximate august australia avionics bajcsy baker based bass benosman brisbane calibrating calibration camera cameras catadioptric chahl complete computational computer conference cooperative coorg creating cvpr dataset deforas devars device digital dimensional directional edition editors eric estimation february fisheye flexible focus folded formation from frontal generation handbook hanson hexagonal hicks high hill hyperboloidal ieee image images imaging immersive incorporating intelligence intelligent international into isbn ishiguro johnson journal july june karmarkar kawanishi kazumasa krishnan kuban large lens lenses life localizing machine magnify manifold master mcgraw measurement michael microsoft mirrors mobile mosaic mosaics moving national navigation nayar november objects omni omnidirectional optics panoramic parts pattern peleg perception peri perspective pose proceedings projection projections properties pyramidal rajasekar range ranging real recognition references reflective report research resolution riseman robots rotate self sensor sensors shum single society srinivasan stereo stryland surfaces system systems szeliski takahito technical technique telepresence teller that theory three tilt transactions true tsuji turkowski using video viewpoint virtual vision volume wide with workshop xiong yachida yagi yamamoto yamazawa zhang zhengyou zimmerman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990508	56	Integrated Face and Gait Recognition From Multiple Views	active alex ambiguity angeles august austria based basu beymer black blanz callari cascia computer conference cvpr essa example expressions face faces facial ferrie from graphics head ieee image images international irfan isbn isidoro local model models morphable motion pages parameterized pattern pentland poggio press proccedings procceedings proceedings recognition recognizing reduce references registration regularization robust sclaroff sequences siggraph society sumit synthesis texture tracking uncertainty using vetter vienna view vision yacoob
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990541	88	Local Feature View Clustering for 3D Object Recognition David G. Lowe Computer Science Department University of British Columbia Vancouver, B.C., V6T 1Z4, Canada lowe@cs.ubc.ca	accurate acknowledgments across adam addition affine allow almost animal another appearance approach approximate arthur artificial available background ballard based basri baumberg being beis bernt better biological boundaries brown cambridge canada carolina centres changes circumstances cluster collins color combination combined comments communication computation computer conference consist constraints continuing cordelia corfu correspondence cost could council crowley current cvpr cybernetics david degree described descriptor descriptors detection developed different dimensional discrimination does done doorn each edge enforcing engineering entirely eric excellence expected explicitly failed feature features field figure formed fort fourier freedompose from further geometric gradients grayvalue greece grimson ground have head high higher hilton hiroshi histograms however human ieee image images improve improved improvement improvementsas incorporate incorporates indexing influenced insensitive institute intelligence intelligent interest interesting internal international invariance invariant invariants isbn james jeff journal july june kang keiji koenderink krystian learning least like linear little local lowe many mass matched matches matching matthew measures mechanisms mikolajczyk model modeling models mohr monkey more most motion multidimensional multiple murase natural nayar near nearest need neighbour networks neurobiology nonlinear numbers numerical object objects only opinion other pami parameters pattern performing points pope pose press probabilistic proceedings properties provide puerto rather ratios recently receptive recipes recognition recovering references regarding reliability reliable replace representation research respect result retrieval rico rigid robotics robustness role ronen rotation scale scaleinvariant schiele schmid sciences search selector separated september shape shimon should show shree side simultaneously single small society solid solution solutions solve south spaces squares stable stephen streams structure structured studies such suggestions suitable supported swain systems szeliski tanaka texture than thank that their they this those three training ture type types ullman under university used useful using valuable vancouver views vision visual well which widely with without work works would
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990518	66	Nine Points of Light: Acquiring Subspaces for Face Recognition under Variable	abscissa acknowledgement adini already also analysis appearance approximation arbitrary athos attributed automation avoided basri because belhumeur believe building certainly changes chen claim code comp compensating completely complicated computer conclusion conditions cone cones conf conference containing corresponding cvpr database david defined delta different differential diffuse dimension dimensional dimensionality direction directional directions discussion each effect effective efficient eigenimages emanating empirical environment epstein equations error exists experiment extended extreme face faces feature flat form framework from function functions future generally generative geometric georghiades gives good hallinan hanrahan harmonic have human ieee illumination image images important individual individuals intelligence intermediate invariants inverse investigation irradiance isbn issues jacobs john july kriegman lambertian large left light lighting like linear machine main many matching maximizing model modeling models moses murase nayar nine norm normals object obtain opposite ordinate other over pami paradoxical partial patt pattern pbmcv photometric plane portion pose possible problem proc proceedings processing prominent provides providing ramamoorthi rate rates rays reasonable recipe recog recognition recognizing reconstruction references reflectance rendering representation represented represents research result results right robotics robust search seemingly shashua show shown siggraph signal simple single small smaller smooth society some sons source sources space spanned steps strauss subspace subspaces successively such suffice sufficient superposition surprising system taking task thank thanks that themes there these this total trans transactions ullman under used using usual variable variation verifying very vision visual ways well what while wiley will with work would yale yuille
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990487	36	Structure and Motion from Uncalibrated Catadioptric Views	address algorithms aloimonos ambiguity antone applications approach automatic based believe bell benosman bruckstein camera cameras case catadioptric commercial computer computing conditions conf conference configurations corresponding current curved cvpr daniilidis decomposable degenerate degree directly editor every faugeras fermul flexible formation four from fundamental geometrically geometry govern head hilton ieee image images increasing independent international intrinsics intuitive isbn island journal june kang labs laws luong make manifold matrices matrix memo minimal mirrors motion multiple number omnidirectional omniview opened originally pages panoramic papadopoulo pattern perspective point points press proceedings provide published questions recognition recovery references relative rendering resolution respectively richardson rotations scene scenes sensor society solutions some sphere springerverlag structure surface technical teller that their this three urban views vision visualization what which with work working workshop year
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990482	31	Investigation of Measures for Grouping by Graph Partitioning Padmanabhan Soundararajan Sudeep Sarkar Computer Science and Engineering University of South Florida, Tampa Email: fg@csee.usf.edu psoundar, sarkar Abstract	algorithm algorithms amir analysis application approach august automata average backgrounds based best between borra case characterization classification clustering cluttered color comparison computations computer conclusions conference convergence correct cuts cvpr data detecting diagonal different domains editor effects eigenvector eigenvectors element elements figure framework friendly from generic graph graphtheoretictechniquesforclusteranal greater grouping hall have hence ieee image implications inadequate intelligence intermediate international introduction isbn journal large leahy learning less level lindenbaum machine malik matrix matula mean measure measures minimization minimum modules most narendra natural normalized optimal organization others pages partition partitioning pattern perceptual performance prentice probably proceedings quantitative range rate recognition references relative respect result ryzin sample sanocki sarkar segmentation shapes significant simple society soundararajan spectral statistically statistics stoer student supervised than thathachar theoretic theory these thornber transaction transactions unifying using value variation view vision volume wagner weiss whereas which while will williams with ysis
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990566	106	Estimating 3D Body Pose using Uncalibrated Cameras	about accurate across action adaptive adjust affine algorithm align analysis angular annals annealed anthropometry appear appendix approach articulated athitsos average badler barron based bayesian between black blake body bone brand bregler cambridge camera cameras capture case cccn chhh clidean coccyx coinciding combining complexity computation compute computer conclusion conference configurations consequence consists constraint constraints construct coordinate coordinates correspondence correspondences cvgip cvpr data davis decomposition degrees degreesonaverage dellaert dempster detection deutscher differences different difficult direct displacements each eccv editor equal equations estimates estimating estimation euclidean expected experts exponential factorization feature features feng figures filtering find first fleet following form formulation frame frames freeman friedman from future gavrila generated geometry given graphical graphics hand hartley hierarchical hinton homography horowitz howe human humans hypothesis iccv identity ieee ijcv image imposing inaccuracy incomplete incremental inference inform initial inputs inthesesequences into introduced invariant invariants isbn isotropically jacobs joint joints jordan journal justifies kakadiaris kanade knowledge known lack laird learning leventon likelihood line linear lines locations main major manual mappings maps match matrix maximum measurement method metric mixtures model models moment moments monocular morris most motion movements multi multiple multivatiate nagel neal neural nips nonrigid normalize note notion novel objects obtain obtained occluding orientation orthographic orthography other outputs over pami paper part particle particular pattern pentland people perona person persons placement plane planes plug point points pose poseaccuracy poses present presented press probabilistic proc proceedings propagation puppetry readily real recognition reconstruction recovery rectifying reference references regression rehg reid rigid rohr rosales rotating rotation rough rourke royal rubin same scale scaling scan sclaroff segments seitz self sequences shadow shape shown sidenbladh sigal since single singular society solution song source sparse special specialized splines ssyi statistical statistics step steps stochastic streams structure subtract such system tail task taylor that theory this thorpe thrun time tomasi towards tracking trained training trans translation twists uncalibrated uncertainty under unified university unknowns used using uuuhhhu value variants varies vectors very video view views virtual vision visual visually wachter were when where which with without work workshop world wusv xzvv yields zisserman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990516	64	Non parametric motion recognition using temporal multiscale Gibbs models R. Fablet and P. Bouthemy	able access acknowledge acknowledgments acquired action activity allows aloimonos analysis applied artificial aspects asusing authors background barbara based bayesian been bergen bertero bobick bold bonet both bouthemy camera capture cbmi chapter characterization cheong class classification coefficients collins comp complete complex compression computer computing conclusion conditional conducted conf conference considered considering content contents copes correct criteria cvgip cvpr data davis demonstrates departement derin designed determining direction discontinuous distribution distributions dublin dynamic each early easy eccv efficient eight energy entropy estimation evaluation experiments exploited exploring extended fablet false farb fermuller fields filters first flow fort frame framework france from functions furthermore geman gibbs gimel handle have heitz horn huang human icip ieee image images indexing information innovation instance intelligence interactions interest isbn italic joint june kluwer knowledge lausanne likelihood line local london markov maximum measurements method mobile model modeling models motion motionrelated movement moving multi multimedia multimodal multiple multiscale mumford natural nelson nonparametric obtained odobez optical oriented over pages pairwise pami parametric paris pattern percentage perception perez perform phil picard pixel poggio polana portilla porto posed presented problems proc proceedings proceeds processing qualitative quite random range recherche recognition references refers regions relaxation relies report reported representation representative resp restoration results retrieval riao rico rigid role royal samples santa satisfactory scale scenes schunck second separation sept sequence sequences shown signed simoncelli since society spatial spatio spatiotemporal statistical statistics stochastic straightforwardly structure supervised support szummer table templates temporal texture textures that theory this tmgm torre toulouse towards trans type unified using video viola vision wavelet were whereas which wide wildes wish with workshop
http://csdl.computer.org/comp/proceedings/cvpr/2001/1272/01/127210307abs.htm	153	Reconstruction of 3-D Figure Motion from 2-D Correspondences	acknowledgements action active address also ambiguities ambiguity analysis andrew animation annealed application approach april archival articulated automate automatic available avoid badler barbara based bayesian believe bernado black blake bobby bodenheimer body both brand bregler cambridge camera capture cardboard carnegie chains cham characters collins colorado computer computing conf conference constraint constraints content cope corfu cues current cvpr dance daniel dariu data david davis denver department deutscher disambiguate dublin dynamic electrical engineering enhance estimating estimation european exclusion explore face facilities figure figures filtering finally fleet footage fort frame framework francisco fransisco freeman from fully further fusion future gavrila generating georgia gesture gleicher goncalves grateful greece hand haritaoglu harwood head hedvig help high hilton historic hodgins hogg horprasert howe human humans hypothesis ieee image improve include includes including inequality information institute intelligence intend inter interfaces interleaving intl ireland james japan jessica jonathan journal june kakadiaris kanade kerkyra killington kinematic kuno larry learned leventon limitation links localization machine making malik manual matthew mellon metaxas michael miura model models monocular morris motion multi multiple multiviewpoint murphy namic nara network neural nicholas nobutaka novel object objects obtaining occlusion other pages parameterized particle pattern pavlovic penetration pentland people perceptual performances perona person plan pose possibility proc proceedings process processing prof program propagation puppetry quality range real recognition reconstruction recovery references refinement registration rehg reid remaining report repurposing retargetting robustness rourke santa school science selection self sequence shadow shannon shimada shirai sidenbladh siggraph singlecamera singularity society specification statistical stochastic stockholm structures such sweden system systems tech technology their thesis this time tool tracking trans unfiltered university ursella useful user using video view vision visual visualizations volume walking ways well werman wide will william with workshop wren yacoob yacov yaser yoshiaki yoshinori
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990490	39	3D-Orientation Signatures with Conic Kernel Filtering for Multiple Motion Analysis	academic achieving adelson america analysis andersson angular application applications approaches approximatation approximate area asada ayer barron based beauchemin bergen besides bigun boston boundaries cells chapter chen cker classification comparisons complex complexity computation computational computer computing conf conference contour controllable convergence cortex current curve cvpr daniilidis deformable department design detection differentiation direct early editors efficient eigenenergy eklundh electrical encoding energy engineering entation estimate estimation european facilitates farid features fields filter filters fleet flow focus folsom freeman frequency from gaussians girosi granlund grzywacz handbook hawaii heeger heitger heydt high higher horn hough huang hypergeometric ieee image independently intelligence international isbn jahne journal junction june kernels kuebler kultanen langley layered layers learning letters level likelihood linkoeping lncs local london long lower machine mase maui maximum means mechanisms method michaelis mixture model modeling models moment more motion moving multi multidimensional multiple networks neural neuronal nonfourier nonparametric objects occluding optical orientation pages patent pattern patterns perception perona peterhans pinter poggio precision press primitive proc proceedings processing puerto quadrature randomized recognising recognition references representation research resolution responses results rico robot robust rosenthaler royal sawhney scale scene segmentation sequence shafer shirai shizawa showed signals similar simoncelli simple simulation smoothness society sommer spatio spatiotemporal spies springer start steerability steerable steering stereo stockholm stopped structure superior surveys sweden temporal texture that theory thesis third this trans transform transparency unified university used using values velocity verlag video vision visual volume wedge weiss when wiklund with xiong york yuille
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990575	112	Decomposed Eigenface for Face Recognition under Various Lighting Conditions Takeshi Shakunaga Kazuma Shigenari Department of Information Technology, Okayama University Okayama-shi 700-8530, Japan {shaku,shige}@chino.it.okayama-u.ac.jp	albedo algorithm also ambiguity analysis application automatic been belhumeur benavente bootstrapping characterization chellappa class classes cognitive comments computer conditions conference corporation crest cvpr database detection determining eigenfaces eigentemplate epstein face faces facial feature fisherfaces from future generative geometry helpful hespanha human ieee ikeuchi illumination illuminationinsensitive images independent integrability integration intelligence international isbn japan jones journal jsps karhunen kind kirby kriegman learning lighting linear loeve machine martinez matching method methods models moghaddam multiple neuroscience object objects ogawa partly pattern pentland photometry poggio press probabilistic proc procedure proceedings program project projection promotion recognition references relief report representation research rftf robust science shading shakunaga shape shapefrom shashua sirovich society specific structure studies subspace supported symmetric technical technology thesis this trans turk under using utilities varying vetter virtual vision visual work yamamoto yuille zhao
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990617	116	A Relationship between Spline-based Deformable Models and Weighted Graphs in Non-rigid Matching	absolute accepted active adaptive aggarwal algorithm algorithms along amit anal analysis anatomical application architecture articulated assignment associated association august badler based belongie below bergevin black blake bookstein boundary brain buhmann building cambridge certain chan character chinese chui coarse coherence communication computation computer computers conf conference correspondence cortical cross cvpr data databases davatzikos decomposition deformable deformations depicted descriptions different digit dimensional distortion drawings dual dubuisson duta dynamic editors elastic elements entries estimating estimation features figure finding finite first fish flow four fourth from generalized generative generic global going gold gradually graduated graph graphical graphs grzywacz hallinan hancock handwritten haralick heart here hierarchical hinton iccv ieee imag image images imaging increased increases indexing inexact information intell intelligence intl invariant ipmi isbn jain jepson jolly july june kimia konen kong lades lange large lavallee learning left leftmost level levine liao line link links local locally mach machine malik malsburg matching mathematical medical metaxas mjolsness modal model models more motion multi neural nonrigid note object objects observational octree optical order original pages pappu parametric patt pattern pelillo pentland philadelphia physics plate point pose press principal prior problems proc proceedings processing puzicha rangarajan recognition reduced references registration relatively representation representations review revow right rightmost rigid sabata same sclaroff second segmented sept sets shape shapes shapiro sharvit show shown siam siddiqi small smoothness society spline splines springer staib statistical step structural structures surface surfaces symmetrybased szeliski tagare technique templates their theory thin third thompson three threshold toga topology trans transactions used using vaillant value values variable verlag vision visual volume vorbruggen wahba wang warped warping warps williams with workshop wurtz yuille zucker
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990509	57	Covariance Scaled Sampling for Monocular 3D Body Tracking Cristian Sminchisescu Bill Triggs GRAVIR, INRIA Rhone-Alpes,^ 655 avenue de l'Europe, 38330 Montbonnot, France {Cristian.Sminchisescu,Bill.Triggs}@inrialpes.fr, www.inrialpes.fr/movi/people/{Sminchisescu,Triggs}	action active adjustment alexandru algorithms analysis anandan anim animation anips annealed applications approach articulated assistance automatic available barr bascle based bayesian bernardo black blake body brand bregler bundle camera capture cardboard cham changes class computer conf conference correspondences cviu cvpr davis deformations dense deutscher discontinuities discussions disparity dynaman dynamics early eccv estimation exponential face fields figure figures filtering fitzgibbon fleet fletcher flow fred freeman from gavrila gesture global gonglaves graphics group hand hanim hartley heap helping hogg howe http huang human humanoid humans hypothesis iccv ieee ijcv image implementation inferring inria interface isard isbn john jojic july june kakadiaris kanade kuch learning leventon line lncs local maccormick malik maps martin mclauchlan media metaxas methods metric model modeling modern monocular morris motion motions movement multi multiple multiview nagel north object objects occluding occlusion optimization outlier parameterized parametric particle partitioned parts pattern pentland people perona persons piecewise plankers point pose posing practical practice prediction primitives probabilistic proceedings processes puppetry quality random rangarajan recognition reconstruction recursive references rehg reid rejection report research rick robust rosales sampling sclaroff selection self sequences shadow shape sidenbladh single singularities singularity sminchisescu smooth society soft solid space specifica specifications springer standard statistics stimulating stochastic survey synthesis taylor tech telea telecollaboration teleconferencing theory through tions toyama tracker tracking triggs turk twists uncalibrated unification ursella using verlag video viewpoint virtual vision visual wachter wiley with without working wormholes wren yacoob
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990661	150	IEEE Conf. on Computer Vision and Pattern Recognition, CVPR'2001	alignement alvey approach april artificial asada august automated automatic automation avoidance based borgne cambridge camera chaumette chesi cipolla combined computer conf conference confluence constraints contours control convergence corke corner cvpr deriche design desirable detector dimensionnal displacement environment epipolar espiau estimation faugeras feedback francisco from function generation geometric geometry graphics hager hand harris hashimoto horaud hosoda hutchinson ieee image images integrating intelligence intelligent international interpolation invariant isbn journal july june korea kriegman littau lncis luong lustman malis manipulators massachusetts matching mechanical mezouar morse motion objects obstacle october oxford pages papanikolopoulos park pattern piecewise planar point position potential press problems proceedings ravani real recognition reconstruction recovery references respect rives robot robotics robots robust rotations sakamoto samson scheme scientific segmentation sensory seoul september series servo servoing singapor singh smooth society space springer stability stephens stereo structure system systems task technique three through time tracking trajectories trajectory trans transactions tutorial uncalibrated university unknown using verlag view virtual visibility vision visual volume voyle with without workshop world zhang
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990452	3	Extended Performance Graphs for Cluster Retrieval D. P. Huijsmans and N. Sebe Leiden Institute of Advanced Computer Science Leiden University, NielsBohrweg 1, 2333CA Leiden, The Netherlands email: fg@liacs.nl huijsman,nicu	about additional advocated approaches architecture base based baumgarten bemmel best better between butterworths case cliffs cluster compare comparison computer conf conference content continually cooper critical cvpr database definition degradation describe development diagonal distributed editors effectiveness embedding englewood equal evaluation expect explicit extensions figures from function fusion generality gives graph graphs growing hall handbook harman hearst heidelberg ideal ieee image infer informatics information investigation isbn letters level lippman logarithmic london made make mars measure measures medical mentioning more much muller multimedia musen need next offers orlando overall overview paper pattern performance porkaew possible precision prentice probabilistic problem proc proceedings proposals query question raghavan range recall recog recognition reducing references refinement relation relevance research retr retrieval rijsbergen salton scope selection showing shown shows sigir similarity simple since size sizes smart society solution specific springer statements storage suggested syst system systems text this total trans trec true values variant vasconcelos versus vision voorhees well what window with
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990473	23	First Order Tensor Voting, and Application to 3-D Scale Analysis	accepted active analysis arbitrary available basic blue boundaries boundary brain brainweb center compendium comput computer conf conference contour corfu coronal cortex cortical curvature cvonline cvpr data dataset definition derose description duchamp epipolar estimation extracted figure fisher from geometry graphics gray greece haralick hoppe html http human ideas ieee imaging information inner institute intell intl isbn kass library lindeberg line lounsbery machine matter mcconnel mcgill medicine medioni mesh meshes models montreal morphology multi multiresolution national neurological noisy outer patt pattern proc proceedings project recognition references refined representation research resolution results robust sagittal scale shape shown siggraph snakes society space stuetzle surface tang terzopoulous three together trans transversal views visible vision white witkin zhuang
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990510	58	Articulated Body Posture Estimation from Multi-Camera Voxel Data Ivana Miki, Mohan Trivedi, Edward Hunter, Pamela Cosman Computer Vision and Robotics Research Lab Department of Electrical and Computer Engineering University of California, San Diego, La Jolla, CA 92093	academic acknowledgement acquisition action active activities activity advantage algorithm algorithms allow also analysis annealed approach appropriate articulated association august austin automation available average background barbara based blake body borovikov bouguet bregler brightness calibration california caltrans camera cameras capture cheung chose compaq computer computers conclusion conclusions conference constancy constrained constraints construction convenient cooperative coordinate corfu could covell cutler cvgip cvpr daimlerchrysler darrell data davis december delamarre demonstrated dense depending depth detection deutscher diego different digital dimensional dimensions dimi directly disparity dissertation distributed division dynamics earlier easily electronics estimates estimation expectation exponential expressive faugeras figure filter filtering forces formulation fortmann frame francisco from gavrila granum greece haritaoglu harville harwood have head highaccuracy higher hilton holler horprasert huang human humans hunter iccv ieee image images implemented improve include inhabited innovation input institute intelligence intelligent interacting interaction international intl introduction isbn island january japan joji journal juan july june kakadiaris kalman kanade kerkyra knowledge kyoto large larger lead learning lenses machine makes making malik manipulation maps march massachusetts mathematical maximization measurable media meeting mentioned metaxas metrology miki model models moeslund monitoring more most motion motions movement multi multilevel multiple murray noise november objects occluding occlusion octree parameters particle partnership parts pattern people perspective physical plan pose posture press proc proceedings program puerto quality qvortrup rahimi rapid rate real recognition recognizing reconstruction references reid relatively reseach research resolution rico robotic robotics robust room santa sastry selection self september sequences shadow shalom shape shelf should simple since size sizes smaller society some sony space springer statistical straightforward subtraction summarization supported surveillance survey system szeliski take technique technology texas that their thesis three through tilt time tracking trans trivedi tsai turk twists unconstrained understanding university used using verlag versatile video view viewpoint views virtual vision visual voxel which with workshop world worlds would wren
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990648	138	Alpha Channel Estimation in High Resolution Images and Image Sequences P. Hillman J. Hannah D. Renshaw Department of Electronics and Electrical Engineering, University of Edinburgh The King's Buildings, Mayfield Road, Edinburgh EH9 3JL Scotland pmh@ee.ed.ac.uk Source and result images at http://www.ee.ed.ac.uk/~pmh/alpha	acknowledgements actor alpha also areas assist assisted attempt august autokey automatic background based because blinn blue bouman bourbakis brinkmann british certain classifier color colour coloured company composite compositing compression computer conference congress content control curless cvpr december degraded detect detection difference differently difficult digital electronic ellis environment epsrc estimation extraction february film first foreground frame future fuzzy generation graphics hillman hint human ieee image images information intelligence interaction investigated isbn june kaufmann live london look machine making matting mitsunaga moghaddamzadeh morgan move moving mpeg natural need number orchard patent pattern performance peter photography possible proceedings processing provided quantization rachael recognition references remove rosin ruzon salesin science screen segmentation sequence shadow siggraph signal similarly smith society spatial states strategies studentship suitable supported system systems technique television that these they third this threshold tomasi totsuka transactions united vision vlahos well werner where will with work world yokoyama zongker
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990619	117	Appearance-Based Object Recognition Using Multiple Views Andrea Selinger and Randal C. Nelson Department of Computer Science University of Rochester Rochester, NY 14627 fg@cs.rochester.edu selinger,nelson	ability absent accidental active adaptive additional adore advantage alarm alignment also appearance approach architecture artificial august available background baek baker ballard based been bernt between bins black borotschnig bruce camps canaria cases caught clutter cluttered color combinations combining computation computer computing conclusions conference conjunction consistency consistent constraint constraints corner correct crowley cvpr delhi demonstrate developed different dimensional disrupted does dramatically draper edelman effect eliminated eliminating errors especially essentially european even every evidence exclude expected experimentally factor false feature features field fighter figure first flynn fourth francisco frequency from further gain gains generated gran graphics greyvalue grouping groups hall have hearn hence hierarchy high histogramming histograms huang hypotheses hypothesis iconic identity ieee image images important imposed improve improvement improves inconsistent independently india initial inspired integration intelligence interesting international into invariants investigated isbn jain james jose journal juan june kanungo keyed kyungim large largely learning learns left level limitations local many matches method missing mohr multidimensional multiple murase nature nayar nelson network neural neurally object objects obtained obtains occur october often ones outweigh over pages paletta palmas parts pattern perceptual performance pinz plateau plateaus plot poggio point pose possible possibly prantl prentice presented proc proceedings produced puerto rate receptive recognition recognize references relations remaining representations research results rico right scale scenes schiele schmid scoring seemore seems selinger setup several shape significantly simple simplicity single society some spain specifically stand structure subject substantial such system systems target tests texture that there this though three through thus tions transinformation understanding upper used using version very view views vision visual weak when where which will with would zooms
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990494	42	Image Magnification Using Level-Set Reconstruction Bryan S. Morse and Duane Schwartzwald Department of Computer Science, Brigham Young University 3361 TMCB, Provo, UT 84602 morse@cs.byu.edu	academic achieve acknowledgments addison adobe again agree algorithms aliasing allebach alvarez amoothing anal analysis anchor anchors andrews annual another anti antialiased applying approach approximation approximations area artifacts arvo average axiomatic ballester based berlin bertalmio better bicubic bilinear blue brown bumpy cambridge case caselles cases characteristics color comments communications company comparable comparing comparison components comput compute computer conclusions conference consider consisting constraints continued contour convenient convolution cook corporation crow cubic curvature cvgip cvpr data deformations densities dependent depends detection different differential diffusion digital dimensional directed disocclusion distributed does downsampling driven each edge edition editor editors effect embedded enhancement equations error even existing exploration explored feiner fifth figure figures filter filters florack flow flows focusing foley followed formulation from fronts frymire funding future gems generated generating geometric geometry gilley glassner graphical graphics green greg gregg haar hall hamilton have heavily helpful hendee however hughes human hunt icip ieee image images impression improve improved information initial inpainting international interplation interpolation isbn isophote iterative jacobi journal kaufmann kimia kimmel kluwer larger lena level lines lions little magnification magnifications magnified malladi maps marginally martin masnou mean measurements medical metaxas method methods might milios minimal minimize mitchell model models monarch more morel morgan morse movies multiscale netravali newell noise noisy nonlinear nonuniformly noted november numer numerically optimal optimization original osher pages parametric park parrots part partial pattern pavicic perception perform performing peter phys pizer point portions possible practice prentice press principles prior problem proceedings process processing properties propogating publishers pursuing reaction recognition reconstruction recovery references removal replication research restoration results romeny samples sampling sapiro sbert schowengerdt schwartzwald second seen selective series sethian shaded shape shapes shocks should siam siggraph significant significantly smaller smoothed smoothing smoothness sochen society space sparse speed springerverlag squared stochastic suggest support supported surfaces synthesis table tannenbaum technique tested thanks that their then this trans transactions tried ullmann under unified university using variations various versions vision visual visually volumetric wallace warping wells wesley whether wilensky with wolberg wong work zucker
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990504	52	Learning Generative Models of Scene Features Robert Sim and Gregory Dudek {simra,dudek}@cim.mcgill.ca Centre for Intelligent Machines McGill University 3480 University St., Montreal, Canada H3A 2A7	addressing agents algorithm american analysis appearance applications approach april arsenin aspects association automation based before belhumeur brightness burgard camera capture change collaboration collins computer condensation conditions conf conference constraints continuity cook corfu cross crowley cvpr dellaert densities density diego direct direction dudek edelman editor environment environments estimation evaluating example expanding exploration farthest features francisco fritz from future greece here higher ieee images intelligence international invariant isbn issues january john jugessur june kerkyra kriegman larger learning learns lighting local localization lowe machine manifold mapping methods milios mobile model models more motion much multi murase must nature nayar nene network object objects order page pages parts pattern perception picard place poggio point pose posed position positioning possible pourraz press probabilistic problems proc proceedings properties raised real recogition recognition recognize references regression rekleitis respect robot robotics robust sample samples scale scene schmid select sensing sept september shashua significantly society solution solutions some sons stable statistical stein structure structured take that thrun tikhonov time tracking training trans translation under using validation viewing vision visual wall washington well what wiley winston with work workshop
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990652	141	Equivalence and Efficiency of Image Alignment Algorithms Simon Baker and Iain Matthews The Robotics Institute, Carnegie Mellon University 5000 Forbes Avenue, Pittsburgh, PA 15213	acknowledgments active adding additive additively affine against albeit algorithm algorithms aligning alignment almost also anandan anonymous appearance application applied approach approximately articulated associated assuming athitsos automatic available backwards baker based because been before belhumeur bergen better black blobs both cascia chance collins composing compositional computer conducted conducting conference constant construction contains contract converge converged convergence converges cootes corresponds currently cvpr decomposition degree dellaert demonstrates demonstration descent described details difference direction directly discussion discussions does eccv edwards efficient eigentracking either error estimation evaluation experiments extension extensive fams fast faster february feedback figure final fitting flexible follow forwards found frame framework frank from full geometry gleicher global gradient group hager hanna have head hessian hierarchical hingorani html http iccv ieee ijcai ijcv illumination image images importance incremental incrementally institute integration international inverse isbe isidoro iterations iterative jepson jianbo jing journal kanade lack like linear local lucas mapped matching matthews model models more mosaics motion much naive naively naval number objects office optimal over pages pami panoramic paper parameter parameters parametric pattern piecewise pixel plot present presented proc proceedings projective publication quite rate rather reach reaches recognition references refs region registration reliable report representation research results reviewers right robotics robust same sclaroff selective semi shum simple slower slowly society space statistical steps stereo studied sundar szeliski table takeo takes taking task taylor technical technique texture thank that their this tracking twice unable under updates updating using varying vedula view vision volume warp warps wide with without wkshp world worse would xiao
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990460	11	A PDE Approach for Measuring Tissue Thickness	acad adams aharon algorithms anal angevine approximations assessment automated avis axis based bayesian birkhauser boston botteron brain buchbinder calculation cambridge cardiol cerebral coll comp computational computer condition conference conservation construction control cores corporation cortex cortical coupled csernansky curvature curves cvpr dale debakey deformable dependent diagrams difference dimensional duncan eberly efficient entropy equation evans evolving extraction february figural figure fischl flow fluid formulations fritsch from fronts function geometrically geometry globally gotto gradient grossman hamilton heart holbrook human ieee imag image images imaging inner interfaces isbn itzhak jacobi jones journal kabani laplace laws left level leveque living louis macdonald magnetic mapp mapping march massie matching material mathematics mceachen measurement measuring mechanics media medical methods metrics miller morse mosby motion myocardial neuroimage nolte numer numerical october optimal orthogonal osher outer pattern photographs physics pizer press prince proc proceedings processing propagation ratnanather recognition references regional resonance riemann schultz science segmentation sethian shape shapes short siam snakes society solvers speed staib surfaces tagare template thickness three tracking trajectories trans transactions tsitsiklis under university using vector ventricular vision wall with zeng zoominvariant
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990611	115	A Region-based MRF Model for Unsupervised Segmentation of Moving Objects in Image Sequences Yaakov Tsaig, Amir Averbuch Department of Computer Science School of Mathematical Sciences Tel-Aviv University, Tel-Aviv 69978, Israel tsaig,amir@post.tau.ac.il	aach acoust adaptive adelson adjacency algorithm anal analysis approach automatic avoiding background based bayesian besag boundaries bouthemy brown camera canny caused chang change choi chou circuits cisst classify clearly coding color commun comput computational computer concentrate conclusions conf conference considering consists criterion cvpr data defined demonstrated detection detects digital discussion distinct dramatically dufaux edge efficient ensures entire estimation even experimental extract fail fast field foreground form from future gelgon generation global graph guided hierarchical highly icassp icip ieee image imaging immersion incorporating increasing information initial instance integrity intell interaction into introduces isbn july june kaup konrad labeling lattice level lippman load localization locating location machine maintained maintaining markov mech memory mester method minneapolis misdetections mobile model moscheni motion moving noise number object objects occlusion only optimized other over pami paper paragios partition partitioned pattern performance porto practice problem proc proceedings processing proposed rainfalling random rather recognition reduce reduced references region regions reliable relying resulting results rico robust robustness same scalability scene scheme science segmentation segmented segments sequence sequences sezan shape should signal simulations smaller smet society soille spaces spatial spatio speech spie static stationary statist statistical stop structural succeeds successfully surely syst systems technique techniques technol technology tekalp temporal than that theory therefore this throughout time tool tracking trans tziritas unsupervised uses using validation various vegas video vincent vision vleeschauwer wang washington watershed watersheds when where which while will with wollborn work
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990644	134	A MRF-based Approach for Real-Time Subway Monitoring Nikos Paragios Visvanathan Ramesh Imaging and Visualization Department Siemens Corporate Research 755 College Road East, Princeton, NJ 08540, USA e-mail: nikos, rameshv @scr.siemens.com	acknowledgments adaptive affine algorithms analysis application approximate austria authors background backrgound based bayesian besag bichsel bombay bouthemy boykov brown buhmann camera canada change characterization chou coetzee collins color colorado communication computer conference connected corfu cuts cvgip cvpr density detection deterministic dirty distributions domaniciu dual energy estimation fast free functions geman gibbs global graph graphics greece greifenhagen greiffenhagen grimson heitz hidden hierarchical huang iarp ieee image images india intelligence international isbn journal kanade labeling like likelihood lipton localization location long machine malik markov methods meyer michael minimization mixture modeling models module motion moving multiscale nagel niemann normalized object objects pages paper paragios pattern perez performance practice problems proceedings processing providing puerto qian ramesh real recognition recovery references region rekers relaxation restoration rico rosin royal scene section segmentation segmenting sequences signal simply situations society some special static statistical statistics stauffer stenger stochastic surveillance system test thank theory this thresholding time topology tracking transaction transactions tziritas understanding using vancouver vated veksler video vienne vision visual volume would zahib
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990527	75	Metric Self Calibration from Screw-Transform Manifolds Russell A. Manning Charles R. Dyer Department of Computer Sciences University of Wisconsin Madison, Wisconsin 53706	absolute additional alternative analysis approach august autocalibration automatic available belgium between bottema calibration camera canonic characteristic computer conf conference constraint constraints construction cross cvpr define derived determinant determined directly disambiguate eahimh editors eigenvalues eimh environments equation european europeanconf expanding expansion experiments faugeras fitzgibbon follows found from fundamental geometries gggmuuui gool hartley himeai hmea holland ieee images imeh information intell intelligence international invariants isbn journal katholieke kinematics koch large leuven lishingcompany luong machine matrix maybank meahh meahhk mecan method metric model modulus moving multiple newyork north note oncomputervision only pages parameterization pattern phdthesis pollefeys proc proceedings projective quadric recognition reconstruc reconstruction references representation representations scale seeeq self sequences sign smile society some springer stratified structure symmetry table that theoretical theory third this tionfromuncalibratedimagesequences trans triggs turn universiteit used useful using vieville views vision what where with workshop yield zisserman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990472	22	Bending Invariant Representations for Surfaces ASI ELAD (ELBAZ) RON KIMMEL Department of Computer Science Technion­Israel Institute of Technology Technion City, Haifa 32000, Israel asi.elad@excite.com	aacademic academy accurate acta algorithm algorithms analysis another application applications applied applying approach artificial automatic barequet based being bending besl between body borg cambridge canonical chapman classical classification closer clustered clustering compared computed computer computing conclusions conf conference contours control cvpr data datasets described developer different differential dimensional dimensions dinosaur distance distances domains dover each effective efficient elad embedding error euclidean example experimental faloutsos fast faugeras ffff figure flattened flattening flexible followed form forms free freeman from game geodesic geometric geometrical geometry giraffe globally graphically groenen hall hand hebert human ieee illustrated illustrates image include indexing input intelegence interactive invariant isbn isometric iterative joint jose journal july justified kimmel kreyszig kruskal lavalee least level machine manifolds mapped marching matching matrix melax method mining modeling modern moment moments most multi multidimensional mutimedia national networks neural next norm november numerica objects only optimal orientation original other output pages paper pattern point points polygon position positioning possible presentation presented press previous primitive problem proc procedure proceedings process propagating proposed purposes recognition recovering reduction reference references representations represents result results review rigid robotics sage scaling scene science section selected separated sethian seventh shapes sharir shown sigmod similar similarity simple slowest society space spider springer squares step steps submmited surface surfaces system szeliski technique techniques tested that theory these this three together towards traditional trajectories trans triangulated tsitsiklis university used useful using versions vision visualisation well were where while will wish with work yields york
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990475	25	A Two-Step Approach to Hallucinating Faces: Global Parametric Model and Local Nonparametric Model Ce Liu Heung-Yeung Shum Chang-Shui Zhang	algorithms analogies analysis appearance application april august automatic baker based benavente bonet break cambridge chaos cognitive computation computer conference cootes curless cvpr database edition efficient efros eigenfaces entropy evaluation face faces fast feret flannery fourth freeman gesture graphics hallucinating hertzmann ieee image images international isbn jacobs journal june kanade learning level liang limits manchester march martinez memory methodology microsoft minimax modeling models moon mosaic multiresolution mumford neural neurosciences november numerical oliver pages pasztor patch pattern pauss pentland philips press principle procedure proceedings quilting real recipes recognition references report research resolution rivzvi salesin sampling second shum siggraph society statistical super synthesis taylor techinical technical teukolsky texture them time transaction transfer turk university vetterling vision
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990492	40	Flexible flow for 3D nonrigid tracking and shape recovery	absolute account achieve action active added after aided algebra algebraically alignment along also analysis ance andv animation appearance applications approach april arbitrary articulated artifacts axis bascle based before best between bicubic birch birchfeld black blake blobs both bottom brand calculated calculations calculus called case certain certainty changes cheeks circuits clean cleft clockwise close closed closing coding columns combining component computations computer conference constraints constructed contour control cootes corner correlation could covari covarianceweighted covell curvature curve cvpr decarlo decompose define definite deformable deformations depth derivation derivative detail determines differencing different differential directionally directions dividsolutions division down each eccv econometrics edwards eigen eigentracking eigenvalues eigenvectors eisert elliptical equally equation equivalent error estimate estimation estimators even everything expression eyebrows face facial factor factors figure find five flexion flexions flow fold forchheimer forehead form frame frames from generalize geometrically geometry girod gives golub graphs greatest hand have head here hilden hopkins horn http human iccv icols identity ieee ijcv image implies improve incorporate indicate inference ingthe instead integration interest interpolation into irows isidoro iteration iwafgr jepson johns joke june kanade known last least left less likelihood loan local location loglike loglikes lucas magnus manuscript matching matrices matrix mean measurement measurements metaxas method middle minimize minimized model models more motion motioncompensated motions mouth must naive naively nats near negahdaripour neudecker norm norms nose note objective objects often opening optical orientation orthonormal pages pami parametric particularly partway pattern peaks permits pighin pixelation point points pose posed positive precise premultiplying preparation present press principal problem problems proc proceedings produces propagating proportional published pulling qvecl raised raising reason recognition recognizing recovered recovers references refined refinement registers removing replace representation resolution results resynthesizing reveals rewritten rigid robotics robust roivainen rotates rotation rotations runs salesin satisfy scale scales scaling sclaroff segment separability separating seqence sequence setting shaded shape show single singleframe skin small society solution solve some spatial specified spherical squares stanford stated statistics subsequent such summands super surface symmetric synthetic systems szeliski taylor technology telling texture that then they this three through time tolerated tolerates tomasi tracked tracker tracking traditionally trans translation translations trouble typically uncertainty unitary unvectorized upper upsampled using variable variables vecrows vector vectorizing very video view views vision volume warp warped weighted well where whereas which widening wiegand wiley will with yacoob zero
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990469	20	On Computing Exact Visual Hulls of Solids Bounded by Smooth Surfaces	about advanced analysis apparent applications astrom automatic based baumgart berger boundaries boyer buehler calibration cambridge carmo cipolla cliffs comp computer computing concept conf conference confluence construction contour contours cross curve curved curves cvgip cvpr differential does dorsey englewood epipolar estimation extremal faugeras from frontier geometric geometry giblin gortler graphics hall hull hulls ieee image images intelligence intrumented isbn jersey koenderink laurentini machine matching matusik mcmillan model modeling models motion multiple nato object occluding octree pattern perception photographs pollard ponce porrill pose prentice press proc proceedings rapid raskar recognition reconstructing reconstruction references report research robert rozenoer sequences shape shlyakhter siggraph silhouette society solid splines stanford stereo sullivan surface surfaces szeliski tech tell teller texture thesis third trans tree triangular understanding university using vaillant views vision visual weiss what workshop zisserman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990495	43	Separation of Diffuse and Specular Reflection in Color Images Stephen Lin and Heung-Yeung Shum Microsoft Research, China stevelin, hshum @microsoft.com	acknowledgements adaptation advantage algorithm allowing america analysis applications approach approaches arbitrary bajcsy baltimore bands based berwick biol bolle book boult breneman brill characteristic chromatic chromaticity collection color colors colour components compt computed computer computing conclusion conditions conf conference constancy constant cook cvgip cvpr daniel described detection diagonal difficult diffuse drew estimating estimation evaluation exists experimental extended fang finish finlayson forsyth from funt generalized geometry give graphics hallikainen handle have healey help highlights ieee ijcv ikeuchi illuminant image images implementation improved independently indexing insensitive intensity intern international isbn jaaskelainen journal kanade klinker kries land light lightness like linear maloney math matte mccann measurement method model modeling models most multi multiple munsell nayar necessary novel number numbers object obtaining optical pages pami parallel parameters parkkinen pattern photometric pixels polarization present presented previous proc proceedings processed recog recognition references reflectance reflectances reflection reflections regionally relation removal requirement research retinex robust sang sato scene schulte segmentation sensor separate separating separation shafer sharpening significant single small society space spectra spectral specular specularities specularity standard straightforwardly suffice sufficient surface surfacereflectance task technique temporal thank that their theory this tominaga torrance trans transformations transforms understanding uniform upon using views vision wandell west when which with wolff work would
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990479	29	Similarity templates for detection and recognition Chris Stauffer Eric Grimson Artificial Intelligence Laboratory Massachusetts Institute of Technology Cambridge, MA 02139	access acknowledgments activity advances analysis annual appear approximate arpa artificial association august automatic based biological boykov cambridge center chris classification classspecific clustering colins color columbus comparing computational computer conditional conf conference content cuts cvpr data densities detection discrete discussions distance distributional energy english example examples experiments face factorization fast fifteenth fort francisco frey from graph grimson hausdorff helpful hierarchical hofmann huizhen human huttenlocher iccv ieee image images information intelligence international invariance isbn jianbo jojic juan june kaufmann klanderman laboratory latent learning leen libraries like linguistics machine malik many marina matrix matsakis media meeting meila miller minimization model morgan muller nature neural nips nonnegative normalized objects occurrences ohio oren osuna pages pami papageorgiou parts pattern patterns pedestrian pereira poggio press probabilistic proc proceedings processing publishers puerto random ratan real recognition references rico rucklidge scene segmentation semantic september sets seung shared sights sinha society solla sounds spectral stauffer sung systems templates thank thanks thesis thomas through tieu time tishby topographic tracking training trans transform transformation transforms uncertainty understanding used using variable veksler video view viola vision walks wavelet words workshop would zabih
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990451	2	Color Constant Ratio Gradients for Image Segmentation and Similarity of Texture Objects Theo Gevers and Arnold W.M. Smeulders Faculty of Science, University of Amsterdam, The Netherlands	access accurate amsterdam analysis anno appl application approaches ballard based belongie bolle books carson classification color comparison components computer conf conference constancy constant content copenhagen cvgip cvpr directed discrimination distributions drew error feature finlayson funt glance gorkani gradient greenspan haralick harwood horowitz iccv icpr ieee image images improved indexing india information international introduction isbn josa journal libraries malik marszalec measures merge multi nayar netherlands nieminen object ojala orientation pami pattern pavlidis photos picard picture pietikainen proc procedure proceedings recog recognition references reflectance reflection retrieval science segmentation sensor separate shafer sharpening society sorting spectral split statistical structural study swain systems taylor texture third transformation university using video vision visual with workshop zenzo
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990457	8	Undoing Paper Curl Distortion Using Applicable Surfaces	aided algorithm apparatus applicable applications approach approximation arrows aumann bending bennis between bezier blake bodduri book bottom cagd cambridge camera carmo center cloth computer conference considerable constructing corporation corrected creasing curl curvature curves cvpr dahelen deformation design detection developable differential digital dimensions distorted distortion document documents duality elber elsevier eurographics fabrication figure fiume flattening general geometric geometries geometry george gotoda graphic graphics grimsdale hall hoscheck ieee iglesias image imaging indicated interface interpolation isbn july kabushiki kaisha kergosien kunii layout less linear lyche mapping mathematical mehaute method methods minolta model modeling nashville number optimal orthoimage paper patches patent pattern piecewise plane platenless point powell prentice press proceedings projection proposed publishing rabut ravani reading recognition reconstruction redont references representation same scale scanning schneider science september showing shumacker shumaker smoothes society spline straightens surface surfaces system technique techniques texture that this university unrolled unrolling using vanderbilt vezien virtual vision visual volume warping with wolberg xerox zisserman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990486	35	Fast and Reliable Color Region Merging inspired by Decision Tree Pruning Richard Nock Universite´ Antilles-Guyane D´ept Scientifique Inter-Facultaire Campus de Schoelcher, B.P. 7209 97275 Schoelcher, Martinique, France rnock@martinique.univ-ag.fr	alfonsin algorithm algorithmic analysis antipolis application approach asian based bayes black bordered bottom brace christophe classification code color comaniciu comments communications comp company compares compensated competition computer concentration conf conference continuity contour cuts cvpr decision discrete editor editors ellis embedding everything fast feature felzenszwalb fields figure find fiorio forms forsyth forthcoming generalization gestalt growing gustedt habib hall harcourt having huttenlocher ieee image images instance intelligence isbn kato kearns laws learnable learning left leung linear lisations local machine malik mansour many markov markoviennes maron mathematics mcdiarmid meer methods modality mode modern most motion multiband multiple multiresol natural near nice normalized obtained optimal ordinateur organization pages part patt pattern perceptual philips ponce pong prentice probabilistic proc proceedings processing procs pruning psis psychology puter ramirez random ratan recog recognition reed references region regions result right robust rosenfeld scene science segmentation sher shows significant snakes society some sophia sourcebook spaces spot springer strategies street supplied table theoretical theory thesis time trans tree unifying union universite using utions valiant valuable variations verlag video vision wertheimer white with woman yuille
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990536	83	AnOpenProbleminMatchingSetsof3DLines	allrunsconvergedtothesameresult almost alsonotethat alwaysndsthebestmatch amethodforregistration andbyonerowandonecolumnmentswithapplicationstomultiple andlocatingof andmatchinga andshowedthatthenonconvergensolutiontoffaugerasandhebert arebothnon asatisfactorysolutionto aswellastheconvergentsolutionpre beslandn bhattacharya cambridge case chenandt closed column computer computervi conference coplanar cvpr dandthe ddynamicsceneanalthecoordinatesystemorigin discussion dlines dlineseg dlinesegmentswithapplicationtopolygonalarc dlinesets dobjects dpolygonalarcs dshapes dtopview extractedlinesegmentsareslightlynon faugeras faugerasandm formcurvesandsurfaces formsolutionofabsoluteorienthealgorithmsarevalidinmatching free fromtheline grimson hebert heisterkampandp horn huang icaa identicationofpartiallyobscuredobjectsintwoandthreedimensionsby ieee ieeetrans ineachcase ingtwosetsoflinesegments inpreparation invarianttothechoiceof isaatplanarobject isbn iterativepointmatchingforregistrationof itsacousticimageis kamgar matching matchingalgorithmmaybeusedforcorrespondence matchingnoisycharacteristiccurves matchingsets matchingsetsofinnitelinesremainsopen matchmeaningless mckay mitpress notethateventhoughthegrid objectmotiones objectrecognitionbycomputer ofinnitelines onmatching opticalsocietyofamer pami parsi parsiandb pattern proceedings proposedin rangedfrom recognition references robotics roboticsres schwartzandm sentedhere setoflinesegmentstoasetofinnitelines severalthousanddierentinitializations sharir sion society springer tationusingquaternions theiterativesolutionconvergesrapidlyand thenumberofiterations therepresentation theresultingmismatchmeasure theroleofgeometricconstraints thisexamplealsodemonstratesthattimation thisrendersthebestysis thus verication verlag vision wediscussedtheproblemofmatchingtwosets wepresentedasolutiontotheproblemsofmatch weranthealgorithmwith weshowonlythe zhang zhangando
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990631	123	Diffusion Tensor Regularization with Constraints Preservation D. Tschumperle´ R. Deriche	academic accepted addison akeley algorithms amer analysis anatomique angeles anisotropic application applications applied approach assoc aubert ballester barbara barlaud bart based bayesian berkeley berlin bertalmio bihan biology blanc blanche blomgren california cambridge case caselles centers chan charbonnier chen cheng circular color comb communication comp complut computational computed computer conference connectivite constrained constructive corners coupled coupling curvature cvpr data department deriche detection deterministic differential diffusion diffusions direction distinct dokl driven ecole ecommunications edge editor embedded emission energy enhancement equations etude examples faisceaux fast fatemi features feraud fiber fibres filtering flat formation forstner framework from geman general geometric geometry granlund graphics gulch haar half harmonic high ieee imag image images imaging implicit incorrectly inpainting inpaintings intelligence intercommission interlaken international introduction invited isbn issue john journal july june kimmel kluwer knutsson kornprobst lbnl lecture level lindeberg location longman machine madrid magnetic malik malladi manifolds mapping maps mareci math mathematics mcclure mcgraw medical medicine methods minimal models morel movies multivalued nationale noise nonlinear notes operator operators orientation orthonormal osher pages paper pardo partial pattern pdes perona photogrammetry photon physica physics points porcupine posed poupon pour precise press probability problems proc proceedings processing publishers puerto quadratic recognition references regularization remote removal report representation research resonance restoration review rico ringach romeny rudin santa sapiro scale scalespace scandinavian science segmentation sensing sept sets shen siggraph signal single sochen society solimini sons soviet space special spectroscopy springer stance statist subde superieure surfaces switzerland tang tannenbaum technical tensor texture theory thesis tikhonov tomography total tract trans transactions tschumperle ucla unconstrained univ university using valued variation variational vector vemuri vision visual vlsm volume volumetric wang weickert wesley wiley with workshop young
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990462	13	Handling Occlusions in Dense Multi-view Stereo	adaptive aggarwal algorithm algorithms andy application applications applied approach approximate april august barnard based baseline bayesian been belhumeur best better binocular birchfield black bobick bombay boykov brown camera carving chai chou collins coloring comp comparison computation computational computer computing conference cooperative correspondence cost cuts cvpr cybernetics dealing december dense depth detectable detection developed dhond diffusion discontinuities dissimilarity dublin dyer early eccv effort emphasis energy enhance epipolar estimating estimation experimental extraction fast filter flow followed formulation fort frames framework francisco freiburg from geiger generalize germany global golland gradients graph greece handling have haynor iccv icip ideas ieee ieeetrans ijcai image images improving initial insensitive interesting intille intl invisible ishikawa iterative january juan july june kalman kanade kang kerkyra kolmogorov kumar kutulakos labeling large layers least line lines locally lucas matching matrix matsuura matthies matting maximum measure microsoft minimization motion multi multiple multiway nakamura nonlinear november occluded occlusion occlusions ohta okutomi only optical optimization other others outlier over pami particular pattern patterns performance photorealistic pixel pixels practice problem proceedings processes progressive puerto rangarajan reasoning recognition reconstrcution reconstruction references regions registration regular rejection report research results review rico robust sampling satoh sawhney scale scene scharstein scheme seems seitz selection september sequences shan shape signal slanted smile society some space springer stage statistics stereo stereopsis stochastic structure subset such surfaces sweep systems szeliski technical technique techniques temporal that theory these tomasi trans transactions transparency true unification used using vancouver veksler versions view visibility vision visionalgorithms visual volumetric voxel will window with work yield zabih zhang zitnick
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990505	53	 	able about acceptance adaptiveappearance adelson affine algorithm alignment along also analysis appearance applications approach articulated back background based being belhumeur biological birchfield black bones brain brightness bushes cambridge camera caused challenging changes clearly close color colour comaniciu computation computer computing conditions conference consistent consistently constraints current cviu cvpr data dempster depicted design despite difficulty discussions disparity dublin dynamic each eccv effect efficient eigentracking elliptical erroneous estimation even fail failure failures features figs figure figures filling filters finally first fleet flow frame frames freeman frey from function future geometry good gradients gray guaranteed hager half have head highlighted hilton histograms html human identified ieee ijcv illumination image include incomplete incorporation independently information intensity into irani isbn jenkin jepson joint kumar laird large last latter layer layers learns level lighting likelihood likelihoods limit locally lock machine matching maximum mean measurement meer memo method misalignment mixture model models modes more morris motion motions moves mpeg multi neural object objects occluded occluding occlusion occur olson optical pami parametric partial pattern patterns peleg perception persist perturb phase pixel pointer possible press probabilistic probabilities proc proceedings projects propagating puerto ramesh real recognition recover references region regions regularization rehg representation results rico rigid robust roughly rousso royal rubin rucklidge runs sawhney scenes search second selected separation sequences shift show shows sidenbladh significant singularity size skin slow small smooth society sources springer stability stable stat stationary steerable stochastic structure subsequently such sufficiently superimposed target template terms tests that then these this through time times tomasi topics toronto total track tracked tracking trans transparency transparent using variable velocity verlag videos viewbased vision weiss when with work would
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990663	152	On the Fundamental Limits of Reconstruction-Based Super-resolution Algorithms Zhouchen Lin Heung-Yeung Shum Microsoft Research, China zhoulin hshum @microsoft.com	achieved adding advances algebra algorithm also american amstrong applications applied area assp assume assumed assumption baker barnard because been below blow blurred borman bose break camera captured carnegie carolina century chinese componentwise comprehensive computation computational compute computer conclude conclusions conditions conference convolved corresponding cubic cvpr dame data deblurring deconvolve decoupled defined device directions disjoint distribute down easy effec effective elad enhancement enlarged estimating estimation experiment experiments extraction face faces factor factors feuer figure finally find first found four fractional frame frames from function fundamental further future gautschi general graphics half hallucinating hardie have head hence high higham highest hilton however huang ieee image images implementation improve improving institute interest interpolation intervals into invariant irani isbn joint kanade larger least lens limit limits linear localness lris magnification mathematical mathematics matrix measured mellon methods minneapolis more moreover most motion multi multiframes must noisy notre number numerical obtain only optimal part parts pattern peking peleg perturbation plane portantly portrait practical press proceedings processing proof providence psfs real recognition reconstruction recovering recursive references region registration report research resolution respectively restoration review robotics sampled schultz sensor sequence sequences several shift should since single size society some south spatial squared squares stevenson sufficient super survey symposia synthetic table task technical that them then theory these time times total trans tsai under undersampled university unless using valenzuela verified vertical vertices video vision volume wasteful where while white will with
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990554	100	A New Signature-Based Method for Efficient 3-D Object Recognition Salvador Ruiz-CorreaÝ, Linda G. Shapiro and Marina Melia ÝDepartmentofElectricalEngineering ÞDepartmentofStatistics University of Washington, Seattle, WA 98105 sruiz@u, shapiro@cs, mmp@stat .washington.edu	algorithm algorithms also analysis andlocatingof appearance applications artificial average besl bialy bootstrap bottom chua clutter cluttered computer conf conference contemp control cosmos cvpr dasgupta density dimensions dobjects dorai efficient efron estimation experiments extensions fall faugerasandm figure form four fraag free from graphics grece harmonic hebert high hilbert ieee image images improvement indexing intelligence international into isbn jackknife jain jarvis john johnson journal kerkyra learning lindenstrauss lipschitz localization machine mappings maps match matching math medioni mesh modeling multivariate murase nayar nearest neighbor nene nition object objects occlusion october other pattern philadephia plans point polygonal positive practice proceedings projection proposed random range rate rates recog recognition references registration representation resampling research resolution respect robotics scenes scheme scott search shown signatures simple society sons space spin ssic stein structural surface table tested theinternationaljournal their theory therepresentation time trans transactions true uncertainity understanding using versus vision visual visualization wiley with yamany york zhang
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990546	92	Robust Point Feature Matching In Projective Space George Q. Chen	accurate adjustment algorithms alvey analytical approach artificial azarbayejani berlin between binford bundle cambridge canny chen combined computation computational computer conf conference constraints corner correspondence cvpr dellaert deriche detecting detection detector eccv edge edges efficient epipolar estimation extraction factorization faugeras feature figure fitzibbon focal frames framework from fujikiy geometry giraudon grayvalue harris hartley iccv ieee ijcv image images intelligence invariants isbn iterative john joint journal kanade kourogiz kurata length leroy level live local luong matching mclauchlan medioni method modern mohr motion multiple nalwa orthography outlier pami parameterization pattern pentland pmai poggio practice problem proc proceedings projective recognition reconstruction recovering recovery recursive references regression retrieval robust rousseeuw sakauey schmid seitz shape society solution sons springer stephens streams structure study synthesis technique tensor theory thorpe through thrun thth tomasi torr torre trifocal triggs uncalibrated under unknown vectors vertex video view vision wiley without zhang zisserman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990468	19	Reconstruction of Specular Surfaces using Polarization Imaging Stefan Rahmann and Nikos Canterakis Institute for Pattern Recognition and Image Processing Computer Science Department, University of Freiburg 79110 Freiburg, Germany fg@informatik.uni-freiburg.de rahmann,canterakis	achieved acknowledgments acquisition additional algorithm also analysis angle apparent arbitrary automation based basis beyerer bhat binford blake boult boundaries brelstaff brooks cambridge camera cipolla clark color components computer conf conference consistency constraints contour contours cvgip cvpr decorrelation deformation depth describe determining deutsche dimensional done editors enhanced even experiment fact factor fang fast form forschungsgemeinschaft free from fukagawa funtions further future general geometric geometry given global graphics healey hierarchical highlight horn hybrid iccv icpr ieee ijcv ikeuchi illumination image images imaging improving inclination incorporating information inspection intelligence intensity international interpolation interpretation intl invisible isbn issue journal kanade kashiwagi kiryati laser layers liang light lighting like local machine measurement method metrology model more motion nayar object objects only optics order oren orientation orientations pages pami pattern perard performance performs photometric polarization polarized polarizers possible presence press proc proceedings processing proof propagation rahmann real recognition reconstruction references reflectance reflection robotics saito sampling sato scale scene schechner sensing separation shading shamir shape simple since society specular specularities specularity sphere spie stereo structured supported surface surfaces symmetrical systems szeliski technique texture that theoretical theory this three trans transactions transparent trucco understanding used using views vision volume wallace well will with wolff work world zheng
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990654	143	On Photometric Aspects of Catadioptric Cameras Takayuki Okatani and Koichiro Deguchi okatani,kodeg @fractal.is.tohoku.ac.jp Graduate School of Information Sciences, Tohoku University Aramaki Aza Aoba 01, Sendai 980-8579, Japan	affects although approach aspects baker becomes between born brooks cambridge camera cameras case catadioptric caused compensate composed comprising computer conference conventional curved curvedness cvim cvpr dependent design designed designs determines difficult directly distortion does falloff fields fisheye formation free from geometric geometry graphics guide have horn ieee image images increase international ipsj irradiance isbn japanese journal journalof known koenderink kogyo lens lenses light major measure mirror modern more much nature nayar notes obtain ogura omnidirectional only optics oren paper pattern pergamon periphery photometric press principles problem problems proceedings processing radiance recognition references relationship remarkable scene serves shading shape sharp shashin shown single society solid specular such sufficiently summary surface takano taken technology that their them theory they this treat trends typical useful using variational view viewpoint vision wide with wolf yagi
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990477	27	Learning Spatially Localized, Parts-Based Representation	academic accuracies accuracy achieved adaptive affects aimed alex algorithm algorithms america amnon analysis appearance application architecture atick based bases basis bell better beymer blind boston bubin characterization comon component components computer concept conditions conference convergence cvpr dashed data david dempster derives develop different dimensional down edge edition eigenfaces eigenspaces example existing experimental face faces factorization faster feature features figure filters first following from fukunaga function further future general give have hawaii herault human ieee image imcomplete includes independent information initial intelligence international introduction investigate isbn january journal june jutten karhunenloeve kirby laird leads learn learning left likelihood lnmf local localized machine march matrix matthew maximum memo methods minimizers minimizing modular moghaddam murase natural nature nayar negative neural neuromimetic number object objectives objects occluding occlusion optical part parts patch patches pattern patterns penev pentland poggio presented press procedure proceedings processing proved rate recognition references representation representations research results right root royal scenes sejnowski separation series seung shashua shown signal sirovich size society solid solution sources square starner statistical subspace such suitable suited synthesis systems tasks terms than that then theory they this tomaso topics transactions turk used using versus viewbased vision visual which will with work
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990657	146	LEARNING REPRESENTATIVE LOCAL FEATURES FOR FACE DETECTION Xiangrong Chen, Lie Gu, Stan Z Li, Hong-Jiang Zhang Microsoft Research China, Beijing, China, 100080 E-mail: liegu@microsoft.com	adaptive advances ahuja algorithm analysis annu appearance application applied approach architecture atick baased baluja barlett bartlett based bell blind boosting burl cambridge cars cereb classes cmuri cogn colmenarez comon compaq component components computational computer concept conference cortex cvpr database decision detection detector discrimination edge effectiveness eigenspaces eurocolt example explanation face faces factorization feature filters fourteenth framework freund frontal general generalization herault hierachical html http huang human iccv ieee independent index information international isbn jones jutten kanade laboratory ladesand learning local localized logothetis macaque machine margin matrix maximum methods modeling modular moghaddam natural nature negative network neural neuromimetic neurosci nips noline object objects oram oren palmer pami papageorgiou part parts pattern penev pentland perceptual perona perrett planar poggio probabilistic proc proceedings processing psychol real recognition references relationships report representation representations research responses robust roth rowley scenes schapire schneiderman sejnowski separation series seung sheinberg signal single snow society sources spatial spatially spie starner statistical structure sung systems technical temporal their theoretic theory thesis time units vasc view viola vision visual voting wachsmuth with yang zhang
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990521	69	JPDAF Based HMM for Real-Time Contour Tracking	ability account acknowledgements adapt adapted allowing allows alpromising also amini analysis approach association background based because birchfield boulet calculate captured changes chesnaud color comfortably comput compute computer conclusion conf conference configuration contour correlated cues cvpr data demonstrate developed different dynamic easily edge efficient elliptical ensure figure filter foreground foundation frames framework from further globally gorithm gradients grant handles have head histograms ieee image incorporate intell intensity inter into jain joint jpdaf jpdafhmm learning line machine material measurements method microsoft model models more national neighboring noise november object obtain occlusion offers only pages paper part partial patt pattern performance physical presented probability problems proc proceedings process programming promising properties proposed real recog recognition references refregier region reported research results robustness same scene science section segmentation september sequences snake society solving sregion stan statistical submitted successfully supplement supported taking targeted test tested testing thanks this time tracking trans transition used using variational various video vision weymouth which work world
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990645	135	Compact Representation of Bidirectional Texture Functions	also analysis appearance arpa automation become block brdf brodatz building class compact complex computer conf conference considered contains cucs cula curet cvpr dana database described determine dimensional dimensionality dimensions direction dover efficient electronic elements experiments features from function geometric ginneken graphics have healey here histogram histograms human ieee illumination image images imaging implications important indeed integration intelligence interactions international irradiation isbn january journal julesz june koenderink kristin learning leung library machine malik matching materials measure measurement methods model modeling modified murase natural nature nayar nene november oana object objects over pages pattern perception planning proceedings quantitative real realistic realworld recognition recognizing redundant references reflectance rendering report represent representations representing robotics sample slam society software spie submitted subsequent such suen sufficiently support surface surfaces synthesis technical techniques textons textural texture textured textures their this three threedimensional transactions understanding used useful using viewing vision visual which with work workshop world
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990571	109	Quantigraphic Imaging: Estimating the camera response and exposures from differently exposed images Steve Mann University of Toronto, Dept. E.C.E. Toronto, Ontario, Canada, M5S 3G4 Abstract	acknowledgments adnan also annual appears applications appropriately august author authors automatically barron bars beauchemin being black boston both calibration camalign cambridge cameras canada charles chart cito combining comparametric comparunroll compositing computer conference corey corley cvpr david debevec december determine determined differentiation differently digital domain donation dynamic each edgerton environments equations estimate estimation experiment experimental exposed exposures extended extending film find first fleet flow from function functions fung germeshausen gives graphics grier ground group high http ieee image imaging individual international isbn issn james joint journal july june known laboratories labs lausanne like line malik manders mann maps march massachusetts method mitsunaga mosaics multiple nayar networks newsletter nortel nserc optical orbit ordered package pages pairs paper parameter pattern performance photographs picard pictures practical presented proc proceedings processing programs projective quantigraphic radiance radiometric range recognition recovered recovering references reflectivity regions related report response result results robust same scene science segmented self serial shifting shown siggraph society statistics steps steve switzerland systems szeliski technical techniques technology test thank them then these this toronto trans transition truth undigital used using values verify verifying video virtual vision washington wearcam well were with would written wyckoff
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990517	65	Rapid Object Detection using a Boosted Cascade of Simple Features	adelson ahuja amit anal analysis anderson andrew application applied area artificial attention baluja bartlett based belongie boosting cars classifiers coarse computational computer conference crow culhane cvpr davis decision design detection detector edgar edward effectiveness eurocolt example explanation face faces features federico filters fine fleuret fourteenth framework freeman freund geman general generalization girosi gooodman greenspan ieee image induction information intell intelligence international invariance isbn itti joint journal kanade koch learning line mach machine machines mapping margin method methods model modeling network neural niebur november nuflo object october oren osuna overcomplete oxford pages papageorgiou patt pattern perona peter poggio press proceedings processing pyramid rakshit rapid recognition references retrieval robert rotation roth rowley saliency scene schapire schneiderman selective shape siggraph snowbased society springer statistical steerable summed sung support tables ters texture theoretic theory tieu training transactions tree tsotsos tuning university vector verlag viewbased viola vision visual volume voting webb wilder william yang yoav york
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990459	10	Frame-Rate Spatial Referencing Based on Invariant Indexing and Alignment with	algorithm algorithms analysis angiogram angiographic annual appear application approach archives artificial augenheilkunde automated avenue barrett based becker beis binford biomed biomedicine blindness bressler choroidal classification clinics coagulation cohen computations computer conference constructing control curved cvpr darpa december degeneration determination digital digitized direct diseases duda editors eichler engineering estimation evaluate evaluation exploitation exploratory extraction feature fine fluorescein following forsyth foundation from functions fundus geometric gitter group hampel hart hierarchical high highorder human ieee image images indexing influence info information intelligence international invariance invariants isbn jacobs january jerath john joint journal june klinische krauss landmark laqua laser lasers levitt linear live location lowe macular madison mapping markow matching medicine meer membranes methods mintz models monahan monatsblatter montage mosaic mosaicing mundy murphy neovascular neovascularization object observations ophthalmology optical optimal order pages pairs pami parameter pattern persistence photocoagulation pillow planar press prevent proc proceedings processing projective puliafito quasi rapid real realtime recent recognition references registering regression related report reports representation research resolution retina retinal review reviews robust roider ronchetti rosenfeld rothwell rousseeuw roysam rylander scene scheduling schomaker september shape shen siam society sons stahel statistics stewart study subretinal surgery synthesis system tanenbaum tech technique theory time tracing tracking trans transformations treatment turner understanding updated using vascular vision vitreous welch wiley without workshop york zimmergaller zisserman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990496	44	Optimal Texture Map Reconstruction from Multiple Views	algorithm aliased appear architecture been blending blur bosse boult buehler captured chiang cohen computer conference corresponding cubic cvpr debevec different distribution effect estimation example figure from generated geometry global gortler have higher highly hybrid ieee image images indicate inputs interpolation isbn kernel less lighter linear local lumigraph malik mcmillan method modeling multiple normalized notice pages part pattern photo photographs pixels proc proceedings real recognition reconstructed reconstruction references rendering resampling resolution respectively result sharper siggraph simple society source spatially super synthetic taylor test texture textures three unstructured used using variant views vision weight weights with zoomed
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990484	33	A Subspace Approach to Layer Extraction Qifa Ke and Takeo Kanade School of Computer Science Carnegie Mellon University fg@cs.cmu.edu ke+,tk	achieves acknowledgements adelson affine after alan algorithm also analysis anandan andareferenceplaner andthehomogeneouscoordinateseeplane anonymous appendix applications approach avidan ayer baker based bayesian bergen between black bouthemy calendar cambridge camera cameras case chen cheng choose cirsysvideo clustering coding coherence collinear color comaniciu comments component computation compute computer conference constraint constraints contain coordinate coordinates corresponding criterion cuts cvpr data decomposition deformations degen denote denotes denoting deriving determined diamondback dimagecoordinates dimensional direction discussions distribution eccv encoding epipolar error estimating estimation exist extracted extraction factorization feature feedback figure flow four frame frames framework free from gaussian gelgon generality geometric geometry given global graph hanna harris harry hartley have hebert helpful here hierarchical hingorani homogeneous homographies iandiare iccv ieee ignore ijcv ijji imag image images incamera incorporating increasing independent indicates induced initial instead integrated into invariance irani isbn isnon isobviousthat isthedirectionofepipolarlines itisobviousthat jepson kanade labeling layer layered layermaps layers least length level like likelihood line lines lipton local loss malik manor markoc martial matching matrix maximum mdkkk mean meer method mixture mmevt mobile mode model models motion moving mris mthatwill mtpppp multi multiple multivariate mundy noise noisy nonerate nonparametric normalized number object objects onto optical optimality order original orthographic orthography other outliers over pair pairs pami paper parallel parallelism parameter parametric part pattern planar plane planes points postprocessing ppkkk pppkkk pppppp ppppppp pppppppppppp ppppppppppppkkk pppr preserved press principal proceedings processing project projection projects proof pxypxyandtppp radius ramesh rank real recognition reconstruction reference references refinement region regions relationship represent representation representing results reviewers rigid robust robustanalysis rppp sawhney seeking segment segmentation segmented segments selected sense sequence sequences shape shapiro shashua shift shown shum simon since singularandexists smoothness society space spaces spatial sprite square stereo streams structure subspace substitute supported system szeliski thank thanks their there therefore this three time tomasi torr torre tracking trans transformation under unified uniquely unit university used using variable vector vectorvtisindependentofthesecondcamera video view vision vtvvvkkkp vvvv wacv wang wehave weiss where whose will window with without work would zabinsky zelnik zisserman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990507	55	Detection and Tracking of Shopping Groups in Stores Ismail Haritaoglu and Myron Flickner IBM Almaden Research Center. San Jose, CA 95120 ismailh@almaden.ibm.com	action activities activity adaptive analysis application asian august azarbayejani background beymer bobick body cmuri collins color comaniciu computer conference cvpr darell darrell david davis detection duggins elgammal enomoto final frame framerate fujiyoshi gavrila gordon grimson guided haritaoglu harville harwood hasegawa horprasert human hydra ieee image institute integrated intelligence isbn january july kanade kiosk konolige larry lipton loughlin machine mean meer mixture model models monitoring movement multiple objects pages parametric pattern pentland people person pfinder proc proceedings ramesh rate real recognition recovery references rehg report rigid robotics robust rosales sclaroff second shadow shift silhouettes smart society stauffer stereo subtraction surveillance survey system technical templates temporal their time tolliver tracking trajectory transaction transactions tsin understanding using video vision visual vsam waters woodfill workshop wren
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990489	38	½	able acknowledgments acquisition actually agapito algebraic algorithm algorithms ambiguities ambiguity analysis angle angles applications approach armstrong around associated auto automatic based been being bookstein boyer broadcast buschmann calculate calibration camera capture carrying cases circle closed comparative compute computer computing conclusion conference conic conics constraints construction contour correspondence cross cuhk currently cvgip cvpr data determine dinosaur does eccv editors estimated estimating estimation euclidean eusipco experiment experiments fact fast faugeras figure fitting fitzgibbon fixed forsyth freiburg from fundamental general generation geometry germany good grant grateful greece hamburg hannover hartley have hayman ieee ijcv image images implementing important invariance isbn jose kneebone least liebowitz like linear lncs london luong machine matrices matrix maybank model modelling models more moving much multiple mundy natural need niem novel novelty number object objects only open oxford paker pami paper parameters partially pattern permission philosophical photographs plane ponce pose possible presented press previous proceedings processing production projective real recognition reconstruction recovery reduced references repeated research resolving results rhodes robust rotating rotation royal scattered sections selfcalibration semple sequence sequences series shape showing shown simple singularities smile society spie splines springer springerverlag stability struction structures sturm substantially such sullivan supplied supported sylvia szeliski table target tensors that their theory this transactions triangular trifocal triplets turn turns uncalibrated university using verified very video views vision wilbur with without workshop workshops yakup zisserman zooming
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990449	0	ImageIndexingwithMixtureHierarchies NunoVasconcelos CompaqComputerCorporation CambridgeResearchLaboratory nuno.vasconcelos@compaq.com Abstractasnotscalingwellwiththedimensionofthefeature	afullyautomated analysisandmachineintelligence andc andplay ballard basedimagequerysystem bayesianmodelsforvisualinforma boston breiman canada chang chapman classicationandregressiontrees classrepresentativewereselectedfordisplayonly colorado colorindexing computer computervision conference content cvpr denver dudaandp ecientcolorhistogramindexingfor equitz flickner friedman gershoandr gray hafner hall hart hierarchicalclusteringtechniquesforimagedatabaseorganizationandsummarization ieee ieeetrans imageretrieval inacmmultime ininternationalconferenceoncomputervision inneuralinformationprocessingsystems inproc inspiemultimediastorageandarchiving internadia internationalconferenceon isbn johnwileyandsons july kluweracademicpress learningmixturehierarchies lippman massachusettsinstituteof massachussetts nalcompression niblack nospecialroleintheretrievaloperationitself olshen onthecomplexityofprobabilistic pages pattern patternclassicationandscene phdthesis proceedings quadraticformdistancefunctions recognition references sawhney smithands society stone swainandd systemsiii technology tionaljournalofcomputervision tionretrieval vancouver vasconcelos vasconcelosanda vectorquantizationandsiganalysis vellaikalandc vision visualseek
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990466	17	Using Robust Methods for Automatic Extraction of Buildings	about aerial against also ameri amount another application april area automatic berlin biwi build building buildings components computer computers computing conclusion conference conic constraints construct correctly curves cvpr depends describing despite detection digital does elevation estimation exsel extract extraction february final first fitting focusing fradkin framework fritsch from fully gabet general generates generation germany giraudon high horowitz ieee image imagery images improvements information initial into isbn isprs journal large leonardis main maitre manual mayer method methods mobile model modeling models multiple munich network noise object objects obtained over pages parameter parametric patches pattern pavlidis performance photogrammetry planar plane planning polygonal polygonalization polyhedrallike presented priori proceedings process quality recognition reconstruction reference references remote renouard report represent representative resolution result results robust roux scene segmentation segments sensing shape show showed shown similar simulation single society stage stages stricker surface survey system technical techniques that then this those transactions tutorial understanding urban uses using validated various vision weak with zhang zone
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990499	47	Tracking Multiple Moving Objects with a Mobile Robot	aaai able academic acknowledgments advanced advantages aerospace algorithm algorithms american angle annual applied approach appropriate arbitrary arrows artificial assigning association automation auxiliary avoid based bayesian been being black blake bootstrap both bottom burgard canada carlo carried clutter commission comparison complexity computer conclusions condensation conditional conference contour contract correspondence cvpr data deal dellaert demonstrate densities density developed direction dynamic each eccv efficient efficiently electronic elfes engineering environment environments estimation eurobot european evaluated evaluation exclusion experimental experiments exponential expressions extracts features february figure filter filtering filters fortmann framework from furthermore gaussian gestures gordon gutmann high hingorani hybrid hypothesis iccv icra ieee illustrate implement implementation implemented increases indicate individual information inherits integrating intelligence intelligent international iros isard isbn january jepson joint journal jpdafs kanazawa keep keeping koller konolige localization maccormick maps matching mathematics measurements meier methods milios mobile monte montreal moravec motion movement moving multiple national networks nonlinear novel number numbers object objects obstacle occlusions office order other outperforms over pages pami paper particle partly pattern person persons pitt pose position possible presented press principle probabilistic problem proc proceedings programme propagation proximity purpose range real recognition references reids reliably represent resolution review robot robotics robots runs russell salmond sample scans science sets shalom shephard simulation sjpdaf smith societies society solve sonar space spaces state statistical still stochastic summary supported systems target technique techniques technology temporal that their they third this thrun track tracked tracking trajectories transactions typical uncertainty under unknown uses using varying vision visual well while wide with work workshop
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990456	7	LDA/SVM Driven Nearest Neighbor Classification	adaptive advances algorithm algorithms along also amari amer analysis approach assoc bellman bottou burges changes chapelle class classification classifications classified classifier classifiers clearly cleveland communications competing computation computer conclusions conditional conference constricted control cvpr dann dept devlin dietterich dimensions discriminant discrimination distance domeniconi duda edition editors effective elongated especially estimates experimental favor feature features fitting flexible friedman fukunaga functions govindaraju gunopulos hand hart hastie homogeneous ianakiev ieee improve improvements improving influence influential information input intelligence international isbn joachims john kaufmann kernel kropatsch large learning leen less local locally location lowe machete machine making measure method methods metric modified modifying more morf morgan most mukherjee multi murshed myles nearest nearestneighbor neighbor neighborhoods networks neural ones optimal over pages paper partitioning pattern peng performance poggio pontil potentially practical presents press princeton probabilities problem problems proceedings processes processing producing programs publishers query quinlan recognition recursive references regression relative relevant report result results rules scale schvlkopf second selection short show signal similar similarity smola society some sons space springerverlag stanford statist statistical statistics stork such summary support svms systems tech tend that theory this tibshirani trans transactions tresp univ university vapnik variable vector vision volume weighted weston when wiley with york
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990543	89	Outlier Detection in Video Sequences under Affine Projection 	abrupt accumulation accurate affine algorithm also analysis angles another answer appears appendix application applications applied approach artificial association automate automated automatic baltimore based because been between bjorck bolles camera cartography changes characterizing columns communications comp comparison computations computer computing concern conclusions conf conference confined conics consensus constraints contained contaminated contribution could criteria cvpr data demonstrate department depths deriche detecting detection determining development different dimensional distances distortion easier easy edition editors effect eliminated eliminating elimination employs engineering epipolar errors estimated estimating european experiments extended fact factorization faugeras features fischler fitting formed frames from function fundamental geometric geometry giving golub good have heyden hole hopkins identified identify ieee image imaged images incorrectly increased intelligence intensity international invariance involved isbn issue iterative jacobs john johns joint journal kahl kanade larger leroy likely linear lines lmeds loan localization locations london long lucas luong machine machinery makes manuscript many matching math mathematical matrix measurement measures method methods missing model more morita motion motionand much multi mundy murray novel numerical observation observations occurred only orthography outlier outliers outlying over oxford pages paradigm pattern planes points possible presented press proc proceedings process projection projective promising puting question random rank real reason recognition reconstruction recovering recovery references registration regression relative research results review robust rousseeuw sample scandinavian scene science sciences segmentation sequences sequential series shape show similarity simultaneous single size society sons spanned sparr stereo still streams structure study sturm subject subspace subspaces such sudden synthetic technique test tested tests than that then there thesis this those through thus tomasi torr track tracking trans triggs truncation uncalibrated uncertainty under university unknown unlikely unpublished used using very video vision volume were when whether wiley window winner winning with worthwhile would zhang zisserman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990532	79	Time-varying Shape Tensors for Scenes with Multiply Moving Points	against aggarwal algebra algebraic also america analysis analyzing anandan applications april associated avidan based bennet bennett between bobick bounds cambridge camera canada carlsson carolina cases changes class classes clemens collins computation computer conference constrained constraints contain contraction correspondences curve cviu cvpr darpa davis depths displayed double dual duality dublin dyer dynamic each eccv elliptical entities european europen exceptions explained faugeras fifth figure first fort frame frames freiburg from fundamental gavrila geometry gesture hand heyden hoffman homography human iccv ieee image images indexing intelligence international interpolating invariants irani ireland isbn jacobs january joint journal july june kanade last like line linearly lior lncs lowest machine manning march matching matrices means measurements model models modulo monocular morphing motion mourrain movement moving multi multiple multipoint nicola note number objects only optical orthographic pages parallax pattern people perspective planar planes plot point points polynomials positioning positions prakash proc proceedings projection projective properties puerto quadrifocal quan recogni recognition recognizing reconstruction reference references relative represent representation representations residual rest review rico rigid rows scene scenes sequence sequences shape shashua show single society some south space springer springerverlag static stockholm stroke stroked structure survey sweden synthesis templates temporal tensor tensors text that their three time tion trajectory transactions triangulation triggs trilinearity uncalibrated using vancouver verlag view views vision visual walking weinshall werman wexler which with wolf workshop
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990512	60	Segmentation and Boundary Detection Using Multiscale Intensity Measurements Eitan Sharon, Achi Brandt, Ronen Basri Dept. of Computer Science and Applied Math The Weizmann Inst. of Science Rehovot, 76100, Israel	academic academinc addison adelson algorithm algorithms analysis anisotropic antonisse application approach august automatic baronti barth based basri belongie berkeley black blur bongiovanni brandt burt carolina casini cgip chan cinque classification clustering code commun compact completion computation computational computer conf conference consensus continuity contour contours control cuts cvgip cviu cvpr data detection diffusion doron driven duda eccv edge elder energies estimation fast favaro from geometry graph growing haar haimes haralick hart heeger hierarchical horowitz http iccv ieee ijcv image information integrating integration isbn jacm john klinger kluwer laplacian leahy leung levialdi lindeberg link liow local lotti machine malik marimont massachusetts meer methods multiresolution multiscale nested normalized optimal pami pattern pavlidis perception perona picture pietikainen press proc proceedings pubs pyramid pyramids ratio reading recognition references region regions review ridge roberto robot robust romeny rosenfeld sapiro scale scene scientific segmentation selection shapiro sharon society sons south space split sprienger structure structured structures tanimoto technique textons texture theoretic theory through traversal tree using variable veksler verlag vision walter wesley wiley with york zhong zucker
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990635	126	Spin Discriminant Analysis (SDA) ­ Using a One-Dimensional Classifier for High Dimensional Classification Problems	additive advances algorithm analysis atick banners bartlett biometrika boosting bose boyd classification classifiers computation computational computer conference cooley crawled cvpr data dept early edinburgh edition editors estimators experiments figure fisher flexible freund friedman from hastie ieee international internet isbn joachims kaufmann kernel large learning logistic maceachern machine making margin methods metric morgan nearest neighbor neural oliver pages pattern practical press proc proceedings processing product recognition redlich references regression report research scale schapire scholkopf schuurmans smola society splines stanford statis statistical statistics technical theory tibshirani tics towards university using view vision visual with workers
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990526	74	Geometric Distributions for Catadioptric Sensor Design R. Andrew Hicks Ronald K. Perline Department of Mathematics and Computer Science Drexel University ahicks, rperline @mcs.drexel.edu	abstract achieved allows analysis angle applications applied appropriate approximate approximately april ascertained bajcsy baker bakstein basis bell block bogner born calculus cambridge camera cameras case catadioptic catadioptric central chahl clearly combining computer conclusions concrete conference confident conic constraints contexts coordinates create cube cvpr definitively describe design designed designs different differential dimensional discussed does each effective engineer equations example figure flat fomin formation four gelfand give given greguss have hicks holmdel hyperboidal ideal ieee ieice image imaging implementation information international introduction isbn january john kawato labeling laboratories lang layed lead leads ljubljana manifolds mapped might mirrors must nalwa nayar need novel omnidirectional onto open optics optimal other padjla pages panoramic paper paradigm partial patent pattern perspective prenticehall press principles proc procedure proceedings projection projections realizations recognition recorgnition rees references reflective report resulting review riemannian robots satisfy scene seemingly sensing sensor sensors sides single slovenian society space specific springer springerverlag srinivasan state states summary surfaces system systems technical television text that then theory this three trans true united university unwarping useful variations verlag vertical viewer viewing viewpoint vision wall well what whether which wide will winter with wolf workshop world would yachida yagi yamazawa
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990548	94	Learning Similarity Measure for Natural Image Retrieval with Relevance Feedback Guo-Dong GuoÝÞÝÝ , Anil K. Jain , Wei-Ying Ma , Hong-Jiang Zhang	actual adaboost almost although analysis application applications approach architecture automatically averaged ballard based better between boosting borko boundary browsing buhmann caivl cases chapter classification classifiers color combine combining comp compared computer computing concepts conclusions conf conference consideration content contentbased contents coorelograms corresponding cortes csvt cvpr cybernetics data databases decision decomposition deliver determined dissimilarity duin early emperical empty euclidean evaluate evaluation examples experimental feature features feedback figueiredo figure filtering finally flower formulae freund furht generalization generally good gupta handbook have histogram http huang iccv ieee image images improves indexing indicate individual information input intelligence interactive internal into isbn jain johansson john jose july kernel kumar laboratory learn learning lots machine makes mallat manipulation manjunath many measure measures mechanism media mehrotra method methods metra minka models mori much multimedia mutiersolution networks neural niblack nips number online open orengo ortega over pami pass patern pattern pentland perception perceptual performance photobook picard powerful practice precision presented press proc proceedings process processing project projects puzicha qbic quantitative queries querying question recall recognition reduce references refinement relevance report representation restricted restriction result results retrieval rubner santini schapire scholkopf science sclaroff search select selection sept shape signal similarity simple slow smeulders society some sons space spaces spie srivd statistical stiles storage stricker such supervised support survey svms swain systems takes tamura technical techniques texture than that them theoretic theory this tieu times tomasi tool tools trans tried ture used using vailaya vapnik vector vectors versus video viola vision visitbjojo visual wavelet weston which wiley with workshop worring wyszecki yamawaki years york zabih zhang
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990524	72	Image-based Modeling and Rendering of Surfaces with Arbitrary BRDFs	above accurate acquiring active adelson after aldinger also although analysis anisotropic apparent applied approach appropriate approximate april arbitrary architecture ashikhmin assigned augmented august avidan azuma background based been before belhumeur bergen between beyond bidirectional both bottom boundary brdf brdfs breakfast bridging calibration canada cast center chai chan cohen common comp composite composited compress computational computed computer conditions conf conference consistent containing cook correctly corresponding course curless cvpr dana darmstadt dataset debevec different discussion distribution dorsey downloaded duchamp duiker dyer dynamic each editors effect effective efficient encompassing entire environments eurographics explore extended extrapolate face factor fast field fields figure final fixed flowers follows fournier frame frames free from function further genc generalization generated generator geometry germnay glass global gortler graphics greenberg grzeszczuk gunawan hand handle handles hanrahan have haven hawkins heng heynen high highlights html http human hybrid ibrl ieee ikeuchi illuminated illuminating illumination illustrated image imagebased images implications information interface interreflections into inverse isbn issues jaun josa july june koenderink koudelka kriegman kutulakos lambert lambertian landy langer large levoy light lighting located lumigraph machine magda malik many maps means measuring method methods microfacetbased middle model modeling modelling models morphing most movshon naturally nayar need needed needs nimeroff note novel object objects observation only onto oren original other outside pacific pages parameterized patt pattern phillips photographs photography photometric pitted pixel pixels plate plenoptic point ponce possible premoze presented press presumably principled proc procedure proceedings processed processing radiance range rate real reality really recog recognition reconstructing reconstruction recover recovering redundant references reflectance reflection region relation render rendering renderings report representation rerendering resulting romanzin rough roughened sagar salesin sampling sarokin sato scale scaled scene scenes science segment seitz selected seoul sequence sequences seven shadowing shadows shape shashua shedding shirley show shows shum siggraph simoncelli since single sketch slice society source sources space sparrow specular still studied stuetzle such sufficient superimpose supporting surface surfaces symposium synthesis synthesize synthetic systems szeliski table taylor tchou technical tensor that them theory there thoroughly tong toronto torrance tracing traditional trans transactions type uiuc unchanged under university unknown used using vallino value vandoorn varieties vase video view viewpoint viewpoints virtual vision visual visualization ward ways well what wheeler which while with without wong wood work yale yields zickler zucker
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990574	111	Generic Model Abstraction from Examples	academic algorithm algorithms alignment analysis appearance approach arbor artificial bajcsy barbara based berlin biederman bombay border brooks case chapter coelho combining computed computer computing conference constraints cormen cubist cvpr deformations descriptions descriptors dickinson efficient elliptical eppstein ettinger european examples felzenszwalb figure finding form forsyth francisco from global graphics graphs greyvalue heller hierarchical hill human huttenlocher ibaraki ieee image images intelligence international interpretations introduction invariant invariants isbn january journal june katoh kluwer large learning leiserson libraries local lowe machine matching mcgraw mine model models mohr mundy murase natural nayar nelson networks norwell object objects october organization pages pami parameterized parametric parts paths pattern pentland perceptual perona pope pose press proceedings processing psychology publishers range recent recognition recognizing recovery references representation research rivest rothwell santa schmid segmentation selinger shape shock shokoufandeh shortest siam siddiqi simple society solid solina structural superquadrics theory transactions ullman understanding unsupervised using variation vision visual volume weber welling winston with zisserman zucker
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990500	48	A Framework for Sensor Planning and Control with Applications to Vision Guided Multi-robot Systems. John R. Spletzer Camillo J. Taylor GRASP Laboratory ­ University of Pennsylvania Philadelphia, PA 19104, USA spletzer, cjtaylor @grasp.cis.upenn.edu Abstract	acquisition active agent agents albuquerque allen alur analysis approaches april architecture august automated automation autonomous autonoumous bajcsy banos barbara becker best blake burgard cambridge collaborative collective computer computing condensation conditional conf conference control cooperating cooperative coordination cvpr december decision density design distributed evasion experimental exploration flannery focus framework francisco fusion games gonzalez guiding hawaii hespanha honolulu ieee integration intelligence intelligent interactive international isard issue issues journal june kabir kotay krotkov latombe lavalle localizationa machine maintaining manipulation mataric maui mobile moors motion moving multi multiple multirobot next numerical observation october parker pattern perception phoenix pito planning press probabilistic problem proc proceeding proceedings propagation pursuit ranging recipes recognition references reliable robot robotics robots santa sastry sensor sensors seventh simmons society soft solution soutter special spletzer stamos stereo strategoes surface symposium systems target teukolsky thrun tracking transactions university vergence vetterling view vision visual with
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990565	105	Calibrated, Registered Images of an Extended Urban Area	absolute accurate acknowledgements acquisition addendum adjacencies algorithms american among analysis annals antipodally antone appear applications approach april architecture argus around automated automatic autonomous award based becker bingham bosse bove calibration camera capture close collins communication compose computational computer conference consortium construct construction contain control controlled coordinate coorg corporation correlation couto cvpr data dataset debevec debitetto decision dense digital directional directories distribution dynamic each earth egomotion elements encoding environment environments eospso establish estimates euclidean example existing exploration extracting extraction eyes facades feron fgdc field finally flexible friedman from full gamut geometric geometry georeferenced geotiff graphics greeve gsfc helicopters hierarchical high highdynamic hill hofmann html http hybrid ieee ijcv illumination image imagery images information infrastructure initial integrating intel interactive into introduction isbn jepson journal june large larson leveson lichtenegger liggett logluv malik manual maps march master mcgraw metric microsoft model modeling models mosaic mosaics multiple muri named nasa national naval networks nions node nsdi observ observing office omni open opengis organization organized orientation orientations overhead pages panoramic part pattern personal photogrammetry photographs plan portal pose position practice preparata presence proc proceedings processing project projective provided quater radiance range rationale recognition reconstruction recovering recovery referenced references refined registration relative remote remotesensing rendering report requiring research resulting rotation rotations rough sample sanders scalable scene scenes semiautomatic sensing sensor series shamos shum siggraph single slama small society spatial sphere spherical spie springer stage statistics status sufficient support symmetric system systems szeliski taylor technical technique teller terraserver test textured theory this tools translation uncalibrated under understanding urban useful verlag vertical view views virtual vision visual visualizing vuong wellenhof wien with wolf workshop world zhang
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990474	24	Support Vector Tracking Shai Avidan MobilEye Vision Technologies, 24 Mishol Hadkalim, Jerusalem, Israel, E-mail: avidan@mobileye.com	academic advances affine allan analysis anandan andrew application approach april articulated axis based bergen bernard black bottou brockhausen burges buxton cambridge care case changing chapter cipolla classification classifier combining comparison computer conf conference cortes cvpr dashed data decision denker denotes detection digit discovery distance drucker editor editors efficient eigentracking estimation european face figure fooled frame frames freund from girosi graph guyon handwritten hanna hierarchical hingorani icml ieee illumination image inference information intensive international intl invariant isbn jackel jepson jerusalem joachims jordan kaufman kearns kernels kluwer knowledge lagendijk learning lecun line lippman machine machines matching mate mateo methods michael mining model monitoring more morgan morik motion much muller multiresolution munich nature networks neural number nuno objects oldenbourg osuna pages pattern press prior proc proceedings processing processings publishers puerto recognition rectangle references region representation result rico roberto robust rules sackinger saitta scholkopf score seen sequence sezan show simard simple simplified smola society solid solla springer stable statistical study support system systems tacker tangent text theory thorsten three tracked tracker tracking training transductive transformation tutorial using vapnik vasconcelos vector verlag viewbased vision volume while with
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990550	96	Enforcing Integrability for Surface Reconstruction Algorithms Using Belief Propagation in Graphical Models	across advanced algorithm algorithms approach approximate arbitrary attenuated august based belief berrou bhuiyan brooks chellappa ciss codes coding communications comp computer conference confidence correcting correctness cryer curvature cvgip cvpr decoding determining direct dupuis empirical enforcing error estimation eustis exact factor fast february field forsythe frankot freeman frey from glavieux good gradient graphical graphs group hall horn icassp iconp ieee images including inference information integrability ishii iwahori january jordan josa journal july koetter kschischang linear linearapproximation local loops loopy mackay markov matrices mean merl method methods models modern multiple munson murphy near network networks neural november object oliensis opteng optimality optimum orientation pages pami pasztor pattern petrovic phase photometric ponce practice prentice probabilities probability proceedings product propagating propagation recognition reconstructing references reflectance sample shading shah shape shoaib society solutions sparse stereo study sumproduct superresolution surface survey theory transactions tsai turbo uniform unwrapping using variational very vision weiss with woodham zhang
http://csdl.computer.org/comp/proceedings/cvpr/2001/1272/01/127211069abs.htm	157	Robust Regression for Data with Multiple Structures	accounted accurate algorithm anal analysis application applications applied approach april aspects associated assumed automated automation band bands bias bilinear biometrika bolles both bounding boyer breakdown bruckstein candidate candidates cartography cases caused center classification clustering college communications comp computational computed computer conference consensus constrained constraint curvature cvpr data december decomposition decreasing defined degree department discarded discontinuities distributions each edition editors elliptical errors estimates estimating estimation estimator estimators every examples exploring extracts fast figure finder fined first fischler fits fitting francisco from functionals ganguly general geometry half hall heteroscedastic high histogram hoaglin homogeneity hough huber huffel ieee imag image imposed increasing industrial inside intell intr introductory isbn isolated issue john johnwiley july june kiryati kittler least leedan leroy level line lines list lmeds mach martin maryland mathematics maximum meer merged miller mintz mirza mlesac model more mosteller multiple multivariate muse nearest neighbordistance noise note optimizing ordered organization outlierdetection outliers palmer paradigm parameter park patt pattern petrou piscataway point points prentice previously probabilistic problem problems procedure procedures proceedings properties random range ransac recognition references regression remaining remove reordered report research rest results reviews robust robustness rosenfeld rousseeuw rutgers sample sampling scale scatter second selected separation sequential shapes sharing shown siam similar society sons special spite squares staircase statist statistical statistics step stewart structures subset successfully surface tables tatsuoka technical technique techniques than that this three through thus time torr total trans transform traversed trends trucco tukey tyler type unbiased under underst understanding uniqueness university used using value values vandewalle variables verri vision weighted were what which width wiley with yohai zamar zisserman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990621	118	A Probabilistic Framework for Graph Clustering Bin Luo, Antonio Robles-Kelly, Andrea Torsello, Richard C. Wilson and Edwin R. Hancock Department of Computer Science University of York York Y01 5DD, UK. atorsell, erh @cs.york.ac.uk	algorithm algorithms analysis annealing application attributed based between boyer buhmann bunke capable case categories change christmas closely cluster clustering clusters combinatorial computed computer computing conference contolled corner correspond cross current cvpr cybernetics data decomposition deforamtions detection deterministic develop dickinson discrete distance distances distinct edit editor eigendecomposition eigenvalues eigenvectors encoding englert entropy evaluation experimental extracting feature fifth form framework from generalized genetic glantz graph graphs groups hancock haralick hierarchical hofmann iapr identify ieee image indexing input intelligence interleaved international isbn iterative jiang journal june kelly kimia kittler laird large learning letters likelihood lncs machine matching matris matrix maximum measure measures median memberships method modelbase modelbases models munger number october operators organization organizing pages pairwise partitioning pattern petrou potential present probabilistic problem proceedings property quantitative random recognition references relational relaxation reminiscent representations rizzi robles same sanfeliu sarkar scene search segen sengupta shape shapes shapiro shock shocks shokoufandeh shows siddiqi sign similarities similarity singular skeletal society spectra spectral steps stereopsis structural structure study systems takes tannenbaum tansactions that this topographic topological torsello towards transactions trees understanding updating using value vector vision visual which wilson with wong workshop zucker
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990488	37	The 3D Line Motion Matrix and Alignment of Line Reconstructions	across aligning aloimonos analysis andreff another applications approach april automatic automation avenue barbara based berlin between california cambridge camera canny canonic collineation comparative computational computer computing conference constructive correspondences cost critical csurka cviu cvpr december demirdjian detection determination determining discrimination displacement edge edition egomotion epipolar errors espiau estimating estimation estimators euclidean expressed faugeras february finding flannery francisco from function future geometries geometry germany give grouping hager hartley horaud huang ieee image images intelligence international investigate isbn january japan journal june known kriegman liebowitz line linear lines location luong machine march matching matrix metric mixed model motion multiple navab nonlinear november numerical obtained orthogonal osaka pages pair pattern perspective piecewise plan planes point portable press priori proceedings projective promising puerto quasi real recipes recognition reconstruct reconstructions rectification references representations results review rico robotics santa scene scenes schmid scientific segmentation segments september sequences servoing sets similar society space spetsakis streams structure study substrate taylor teukolsky time toyama tracking transactions uncertainty understanding university using vetterling vieville view views ville vision visual when will without work zhang zisserman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990523	71	Relighting with the Reflected Irradiance Field: Representation, Sampling and Reconstruction	acquiring adaptation adelson anisotropic annual applications approximation april assumption august available based belhumeur bergen bishop bombay bouguet bouguetj calib calibration caltech camera carpenter chai chan chapter cohen computational computer conditions conference constant controllable cook cvpr debevec depth dimensional display distributed duiker dynamic early eighth elements etienne eurographics face fairchild faugeras ferwerda field france from function functions generalization geometric global gortler graphics greenberg grzeszczuk hanrahan hawkins heng high horn http human iccv ieee ikeuchi illumination image imagebased images india interactive inverse isbn january july june kriegman krueger lafortune lambert layered levoy light lighting linear lumigraph malik maps march masking matlab mcmillan measuring model modeling models multiscale nayar needed number object observation olivier oren pages panoramas pattanaik pattern pellacini photographs plenoptic porter possible press proceedings processing psychophysically radiance range real realistic recognition recovering references reflectance reflection relighting rendering robot sagar saint samples sampling sarokin sato scattering scene scenes series shape shirley shum siggraph society spatial structure subsurface surfaces synthesis system szeliski tchou three tong toolbox torrance toward tracing under variation viewpoint vision visual ward what wheeler with wong workshop zhang
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990547	93	Shape contexts enable efficient retrieval of similar shapes	accuracy active affine agreement algorithm amit appearance application applied approximate approximation architecture automation based bayesian belongie bengio biederman bookstein bottou buhmann burges canada carlsson categories chan chui classifiers closed coil columbia communication complexity components computer computers computing conf conference contemp contour cooper cootes correspondence cucs curse curves cvpr databases decomposition deformations descriptors detection dimensionality distortion document dynamic eigenspace eighth experimental extensions face familiarity features fize fourier from gavrila geman gradientbased graham grouping haffner hagedoorn hausdorff hebert hilbert human huttenlocher ieee image improving indexing induction indyk international into invariant isbn jebara johnson joint journal july june kimia konen kopf lades lamdan lange learning lecun library lilien lindenstrauss link lipshitz lncs machines malik malsburg mapping march matching math measure memory merlot model models moghaddam motwani murase name nature nayar nearest neighbors nene nips norms november object objects olson order oriented pages pami pattern pentland philomin pictures plane plate point points principal proc proceedings processing psychological psychology puzicha rangarajan real recognition recognizing references removing report representation review rigid robotics roskies schol schwartz sept shape shapes sharvit smart snodgrass society space speed splines springer standardized state structure support symmetry symposium system taylor technical their theory thin thorpe time towards training trans tree understanding univ using utrecht vancouver vanderwart vector vehicles veltkamp view vision visual volume vorbruggen warps wilder wolfson wurtz zahn
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990515	63	Tracking and Modeling Non-Rigid Objects with Rank Constraints Lorenzo Torresani Danny B. Yang Eugene J. Alexander Christoph Bregler	absolute active america analysis anandan animation application artif artificial automatic bascle based bergen biermann black blake blanz body bouget bregler closed closedform communication computer conf conference constraints contour contours costeira cvpr data decarlo deformable driven eccv error estimation expression expressions face faces facial factorization features fleet flexible flow form foundation frame from funded gesturerecognition gestures gokturk good grimm grzeszocuk guenter hand hanna hecker hertzmann hierarchical hildne hingorani horn human iccv ieee image images intell intelligence international interpretation irani isard iterative jepson joint journal kanade kass lanitis learning lischinski local lucas making malvar matrices metaxas method model models monocular morphable motion motions multi national negahdaripour optical orientation orthography orthonormal pages parameterized parametric part pattern photographs pighin pose private proc proceedings program quaternions realistic recognition recognizing recovering references registration residual resynthesizing reynard rigid ruderman salesin science separability shape siggraph snakes society solution stanford stereo streams subspace synthesis synthesizing szeliski technique terzopoulus through tomasi track tracking under unit using vetter vision visual with witkin wood workshop yacoob
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990481	30	Efficient Spatiotemporal Grouping Using the Nystrom Method ¨ Charless Fowlkes½¾ and Jitendra Malik½	able acad academic achieve acknowledge acknowledgements across adelson algorithm alice analysis andrew anwendungen application approach approximate approximation artificial auflosung august austin authors automatically baker based belong belongie bergen berkeley between bezdek bipartite blowing bolles both boundary boyer breakout buhmann cambridge change clarendon clarke clearly clustering coherence coherent color column commentationes component computation computational computer conclusion conference contour correspondence course cues cuts cvpr data delves demonstrated dept determining development dhillon dimensions directly discussions documents domain donato doretto during dynamic each earen eccv edition editors efficiently eigenvalue eigenvalues eigenvectors emergence encouraging energy epipolar equations equivalent essa estimating estimation example extended facilitate factorization fagerstrom feature fellowship figure find firetruck flannery flow flower form four frame frames framework freeman from further garden gating gdalyahu general gepshtein gianluca graph greedy group grouping groups hall have helpful high hofmann hope iccv icml ieee image images importance important incorporating incorporation indicate indicated individual input integral integralgleichungen intensity intervening introduced isbn johansen journal june kahan kernel kluwer kubovy large layers leap learning leaves left leung like likely likewise location machine machines makes making malik many marimont matrix measures mechanisms medical method methods micro mixture models mohamed more motion muller ncut needed neural nips nonlinear nonparametric normalized note number numerical nystrom objects opposite optical order organization original oxford pages pami parametric part partitioning pattern perception perceptual perona physicomathematica physics pixel pixels plane pocv possible potentialtheorie praktische presented press problem problems proc proceedings provides publishers puzicha quality quantitative randwertaufgaben recipies recognition references region remaining report represents results retrieval review rows salesin same sarkar schodl scholk seeger segment segmentation segments sequence sequences session should shown shows sides siggraph similarity simultaneously slices smola smoothness soatto society solve sophisticated space sparse spatial spatio spatiotemporal spectral speed static stochastic strong structure such supported surface sweep systems szeliski technical technique techniques temporal teukolsky texture textures that there this through time tracking trans treatment tree typical uber unified university unsupervised using valuable variety version very vetterling video vision visual volume voxels weinshall weiss werman which will william williams wind with within words work would zheng
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990544	90	Spatial Information in Multiresolution Histograms	adelson analysis annals annual approach approaches april bach ballard barron based bergen blackman block bonet brazilian browsing burt carlen carlotto central code cohen color colorado combining communications compact comparing computation computational computer computing conference connections content control convolution copenhagen correlograms correspondence crowley cvpr dana databases denmark denver diku discrimination distance earth engine entropy evidences experimental factor field filtering fisher flexible framework fuler gaussian geometry ginneken gives graphics greece griffin grossberg guibas gupta hadjidemetriou hampapur haralick heeger histogram histogramming histograms horowitz huang humphrey iccv ieee ijcai ijcv image images imprecision indexing inequalities inequality information inspired integrated isbn jain joint jose journal july kerkyra koenderink kumar laplacian limit management march mathematical measure measures miller mishima mitra model mover multidimensional multimedia multiplicative multiresolution nayar neural neurally niblack nonextensive object oomes open orderlessimages orengo pages pami parsing pass pattern physica physics plastino powers preserving probability proc procedure proceedings production project pyramid qbic quantities querying real receptive recognition references reflectance retrieval sampling satisfied scale schiele search sets shannon shape siggraph similarity smoliar snoeren society soffet solution some space spaceapproach spaces spie sporring stam statistical statistics storage stricker structural summation surfaces swain synthesis systems tanaka target texture theorem theorems theoretical theory thermostatistics transactions transformation transformations tsallis under using variable video viii viola virage vision visual volume watanabe weickert which without witkin workshop world zabih zhang
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990528	76	Convex and Non-convex Illuminant Constraints for Dichromatic Colour Constancy	acknowledgements akkiraju algorithms alpha authors based best both colour colours common comparable computational computer conference constancy constraints convex convexconstraints cvpr date definition derivation dichromatic edelsbrunner enforcing error estimation experimental facello fall geometry greatly half hewlett ieee illuminant image improved incorporated international intersection introduce isbn large least mathematical model more muck packard pages pattern performance planes proceedings provides recognition reduces references reflectance restricting results satisfy shapes show society software solution solve squares starting supporting than thank that these this varela vision wish with within work workshop
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990640	130	Using an ICA Representation of High Dimensional Data for Object Recognition	academic accurate achieves adapted adaptive advances affine algorithm allow amari analysis appearance approach appropriate architecture assumption ballard barcelona based bauckhage bayesian bell blake blamed blind boston both cardoso carnegie cichocki classification classifier classifiers code color comon comparably comparison component components computation computer concept conclusions conference correspondence costly criterium crowley cvpr data databases datasets deconvolution dempster denoising densities density detection detectors difference dimensional disadvantage discriminative distribution duda eccv edition employed estimation evaluation even fast feature features fields first fixed focuses fogarty frey fukunaga future gaussian global goal guillamet hart herrault high histograms hollandstyle hyvar hyvari iccv icpr ieee implement implementation incomplete independence independent indexing inen infomax information interest international introduced introduction invariance isbn john journal jutten kanade kaufmann kernels keypoints laird latter learning leaving letter letters likelihood local localization machine main march mateo maximization maximum mellon merz method methods mohr morgan motivated multidimensional nature nearest neighbor neighborhoods neural neuromimetic nongaussian nonparametric nonprobabilistic object occlusion only outperforms pages paper part partial pattern pentland performance performed performs point precision presented press prior probabilistic problem proc proceedings processing programs quinlan receptive recognition references relevant report repository representation requires research resulting rotation royal rubin sample schiele schmid second sejnowski selected selection separation shown shrinkage signal simple slate society soft sons source sources spaces sparse springer stage standard statistical stork straightforward structure such supergaussianity swain tech technical technique techniques tensor then theory this those through tomasi tracking training transformation university using vapnik verdie verlag versus vision visual vitria wiley within without yang york
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990464	15	Dynamic Depth Recovery from Multiple Synchronized Video Streams1	adiv aggarwal algorithm analysis anandan application approach approximate august baker based bayesian belhumeur bergen berlin binocular bombay boykov calibration camera capture carceroni carolina carving chellappa collins color column combining computer conf conference constrained correspondence criterion cuts cvpr cybernetics december depth determining dhond dimensional direct dynamic eccv edge energy estimate estimated estimation european fast faugeras figure flexible flow formulation from generated global graph hanna heidelberg hiearchical hingorani iccv ieee image images india integrated intelligence interpolation isbn january joint journal july june kambhamettu kanade kutulakos lecture left lhuillier machine maps march matching maximum minimization model models motion moving multi multiview neil noisy notes objects okamoto optical original pattern problem proc proceedings progressive proposed quan rander recognition recovery reference references reflectance results review right rigid sampling sawhney scene scheme science segmentation seitz sept sequence sequences several shan shape shows society south space springer stereo stereopsis streams structure surfel system szeliski technique theory three transactions triangulation uniqueness using vedula veksler verlag view vision wacv workshop young zabih zhang
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990553	99	AnalysisandSynthesisofHumanFaceswithPoseVariationsbya ParametricPiecewiseLinearSubspaceMethod	ahuja akamatsu algorithms andc andd ande andk andmachineintelligence andp andr andt arepresentationfor articialintelligencelaboratory atkeson basedimageanalysisandsynthesis beymer beymerandt bienenstock bishop bochum carcassoni ceedingsoftheinternationalconferenceonautomatic choi computer conference conferenceonautomaticfaceandgesturerecognition constructiveincremental cvpr deguchi dfacialposeand dfacialposeusingtheemalgorithm doursat estimating example exampleview facedetectionusingmixturesoflinearsubspaces facerecognition facerecognitionbyelasticbunchgraph facerecognitionfromone faceswithbetterfeaturecorrespondenceforimagefaceandgesturerecognition fellous ferenceonautomaticfaceandgesturerecognition feretevaluationmethodologyforface fromtheorytoapplications gazepointestimationusingarobustreal geman generationusingpca genericmodel grenoble hancock headposedeterminationfromoneimageusinga heinzmannanda ieee ieeetransactionsonpatternanalysis ieice imagesynthesisfromasingleexampleimage inpro inproceedingsoftheinternational inproceedingsoftheinternationalcon inproceedingsoftheinternationalconferenceofarticialneuralnetworks isbn isonoands kriegman kruger learningfromonlylocalinformation linearobjectclassesand malsburg matching maurer medialaboratory moon nara networksandthebias neural neuralcomputation neuralcomputing neuralnetworksforpatternrecognition newyork oxforduniversitypress pages paradigm pattern phillips poggio potzsch proceedings rauss recognition reconstructionfromgraphslabeledwith references responsesofgaborlters rizvi schaalandc shashua shimizu society springer technicalreport technicalreporthip timetracking transactionsonpatternanalysisandmachineintelligence variancedilemma verlag vetterandt vision vonder wiskott yang zelinsky zhang
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990519	67	Illumination Invariant Face Recognition Using Thermal Infrared Imagery	access ality alternative analysis answer appears applications based beat been beyond biometric cannot cases challenging classification classifiers collect collection collections combination common compared compelling computer conditions conference conjecture contain continue continuing cost counterpart cover current curse cvpr data database declining definition deployment development dimensional dimensionality disadvantage effort enough environmental especially essence expand experiments extend face facilities fact further future great have head high higher hilton history holds however identification ieee imaged imagery important increases indoor infrared instance intend intrinsic investigation isbn just last local lower lwir main methods modality more much nature necessary neither note obtained ones operating order over pattern performance perhaps poor possible potential prepared present price prices proceedings prokoski promise provides question rather reason reasonable recognition references representative results rich scenarios searching secure sensors should significantly since situations society source sources special spectrum statistical status steadily study subjects substantiate such systems than that their thermal they this though thus time under using variance very visible vision visitor volume were which while will with workshop would years
http://csdl.computer.org/comp/proceedings/cvpr/2001/1272/01/127210805abs.htm	155	On the Individuality of Fingerprints	able about acceptance account acquisition action additional additionally addressing adversary alpha also american among analysis answered answers anthropology approximation association associations attributes authentication automatic available barcelona based bearing been behavioral being believe between bias biometric biometrics bolle both bounds brute byron calculation capacity cappelli case challenged character characteristics characterization class clear closer cole community comparison compete competition completely complex complexity composition computer concepts conclusions conditions conference confidence confounding considering correspondence counts court criminal critical cummins cvpr data databases daubert dependencies deployed dermatoglyphics detailed detection determining development different difficult dimensional directly discriminable distinctiveness distribution district does dover eastern engineering enormous entails entropy epidermal error errors establish estimates evaluation eventually expected expert extension face factors false features finger fingerprint fingerprints first force forensic form formalize formula foundations frequencies from fundamental future genuine government guess guessing hacker have hearcourt here high higher hong http human iapr identification identity ieee ignores illformulated image impostor impostors include including increase increasing indicates individual individuality information inherent input insight interaction interactions international into introduction invest involves isbn issue issues jain journal known lead legitimate like likelihood limited lower maio maltoni manual matcher matchers matching matchof mates mathematical mature memorandum merrell midlo minutiae mitchell modalities model modeling models more most nature negative nevertheless number numerical objective objectively obtained obviously occurrence only order osterburg other palms pankanti parthasarathy password pattern patterns pennsylvania performance permanent personal pharmaceuticals physical physiological poor population pore post postd practical predicted predicting present previous probabilities probability probably problem problems proc proceedings process proposed publications quality quantification quantitative quantitatively question questions raghavan rate rates realistic reasons recognition references reinforced related relatively representation requiring research ridge ruled samples scalability scales scars sciences sclove sense sensitive setting signal significant significantly simple simplicity social society soles someone space spain spite spurious standards states statistical stiles stoney stringent strongest study such support suspect sweat system systems table techniques than that theoretic these they this thornton time topic total trauring types understanding unfortunately unibo united upper usaoedpa using veridicom verification very vision wayman weak well wellunderstood what which while whorld will with would york
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990471	21	Robust Crease Detection and Curvature Estimation of Piecewise Smooth Surfaces from Triangle Mesh Approximations Using Normal Voting D. L. Page, A. Koschan, Y. Sun, J. Paik, and M. A. Abidi Imaging, Robotics, and Intelligent Systems Laboratory The University of Tennessee Knoxville, TN 37996 page@iristown.engr.utk.edu	aaai acadmeny acknowledgments across again algorithm allows amsterdam analysis approximation based both boundary charge chen comparison complex computational computer computing conclusions conference constant control crease curvature curvatures cvpr cylinder data defined demonstrated description detection direction discontinuities distributions edge electrical elsevier error estimating estimation european extension extraction feature fifth figure flynn framework from gaussian geodesic grant graphics graphs greece grouping hagen hans have heinz highlight ieee improvement improvements independent information intelligence international internationaljournalofshapemodeling intrinsic introduce introduced isbn italy iterative jain journal kerkyra kimmel kmax kobbelt level levine ligure lines machine manifolds margherita martin medioni menlo mesh modeling morphological national neighborhood noise noisy normal normals nssa number offers only onmmnis operators orientation original over pages parameter park part paths pattern perry peter piecewise plane polyhedral presented press principal proceedings program properties range ranger real recognition references reliable representation research results robotics robust rossl santa scanner schmitt schreiber sciences segmentation sept sethian seventh shape show shows similar simulated simulation size smart smooth society some specified sphere stable supported surface surfaces symposium synthetic tacom tang taubin tensor that these thesing this tilted timewherethenumberofvertices tolerable transactions triangle triangulated triangulation truth under university user using variance vertices vision vmin volume voting where with work workshop
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990659	148	Optimal Adaptive Learning for Image Retrieval	access adaptive adelson algorithm america application architecture archiving asilomar assist august austria basalaj based bayesian benchmarking bergen book boosting boston brodley browsing cbms chang circuits class communication company computationally compute computer computers conf conference content conv current cvpr databases daubechies decision digital directions discriminant does early edge editors edward efficient eigen embedding estimation examples experiments expo extraction faloutsos feature features feedback filling filter france gary gudivada guest hall haykin hierarchical hill hoff hong huang human icip ieee image implementation information interaction interactive introduction isbn ishikawa issue issues japan jijay journal kobe learning lecture lectures libraries library lippman lyon macarthur machines magazine maps march marchionini mars mcgill mcgraw mehrotra metric miller mindreader minka models modern motion multi multimedia multiple notes novel november omohundro open optical optimizing organization pami papathomas pattern peng picard pichunter prentice probabilistic proc proceedings processing promising psychophysical query raghavan recognition references relevance relevant representation retrieval river rodden saddle salton santini sciences second series shyu siam signals similarity simon sinclair smeulders society special storage subramanya support swets switching system systems techniques theory thirty through tian tieu trans trees update upper user using vasconcelos vector venkat video vienna viola vision visual vldb water wavelets weights weng wescon widrow with wood workshop worring years yianilos york zhang zhou
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990493	41	Instant Dehazing of Images Using Polarization Yoav Y. Schechner, Srinivasa G. Narasimhan and Shree K. Nayar	actually addison adelson algorithms analysis angle appendix applications approach architectures atmospheric attenuation averaging ballard based beams bellingham bohren boult brooks cambridge camera chandrasekhar chromatic color colorchecker comp compensate components comprised computer computing conditions conference containing contrast cost cozman cvpr decorrelation degradation degree depth detector different digital dover dynamic effects engineering enhanced enhancement estimated exposures ezra fang farid framework from fusion gated glassner grewe guidance hecht high highlight iccv icip ieee ikeuchi illuminator imag image images imaging improving inclination incoherent independent invisible isbn josa kashiwagi kaufmann keying kiryati kopeika krotkov landing laser layers light lighting linearized macbeth maximum measurement model monochrome morgan mounted multi multiple narasimhan nature nayar nostrand oakley objects obtained onnen optics orientations partially pattern pencipowski photographs physical polarization polarizationbased polarized polarizer poor press princeton principles proc proceedings processing quality radiative radiometric range recognition reduction references reflection reflections reflector removing response resultant saito satherley sato scattering scenes schechner segmentation semi sensor sensory separate separating separation shamir shown shurcliff shutov signal society solomon spie statistical surface sweet synthesis synthetic system taken technol these this through tiana trans transfer transparent understanding university using vehicle videopolarimeters visibility vision weather weighted were wesley with wolff york
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990551	97	Automated Cartridge Identification for Firearm Authentication Jie Zhou, Le-ping Xin, Da-shan Gao, Chang-shui Zhang and David Zhang+ Department of Automation, Tsinghua University, Beijing 100084, China	abdeljaoued accuracy acknowledge acknowledgements adaptive advances algorithm algorithms anisotropic another astronautics authors automated automatic ballistic ballistics based berlin better bijhold blobs boston breach burger burges carnahan carolina carried cartridge cartridges case cases china chinese chosen circle comparative compare compared comparison computer computing conclusions conference correlation cross cvpr data database degree denote detection different ding distance distributed each edited effectiveness efficient ejector enforcement england evidence existing experiment experimental extraction face feature features figure filtering finding firearms firing firmin flow forensic foundation from fusion gaussian geradts germany gradient grants guarantee have head hermsen hilton hough human ibis identification identify identity ieee illingworth image images imaging impression including infrared integrate integrated intelligent intergration international investigation isbn iuen joachims journal june kernel kernels kinds kittler large learning making marks matching mayoraz method methods michael minor model much multisensor natural networks neural north novel obvious optical paper pattern peng performance person polynomial position practical presented press prince princen proceedings processing projectile proved ratio recognition references represent representation respectively results rotated sanderstead scale science security segmentation several shape shapes simple simplified single sklansky smith smola snakes society specimens speech spie springer statistical structure study support system systems table techniques technologies technology test texture that theory this thompson threshold trans transform under underwood used using vapnik varshney vector verification verlag vision watson when which wiley wish with yacoub yang york zhang zhou
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990662	151	A Model-Based Road Sign Identification System	about above accounts accumulate accv ahmad akatsuka allow also analysis annealing applic application automated automatically automation automative autonomous azami based before being bergevin betke blancard both camera carried central characterization chevron circular claimed classes closer color colors colour compare comput computer computing conclusion conference considered constitute containing content contribution control cues current cvpr database daylight decision description detection developped different dimensional dimensions distributed distribution each earlier encoded environment estable evidence fast faugeras figure flexibil frame frames from function further gavrila generated geometrical great grouping guidance highway huang iccv identification ieee image images imai incorporating increasing information infrastruct input intell intelligent interface invariance invariant isbn josa judd kehtarnavaz knowledge labelling larger largest lauzier learning lies limited localization lower mach main making makris matching math means memo mentioned method micheli minimal model modelling modular moving must newly noisy novel object obtained offers outdoor over parameters pattern perceptual performances performed philomin physics piccioli plays point present presented press priese prince problem proc procedure proceedings processing provides pursuit quasi rates real recognition recognize references relying reported representative results ritter road robust role route saint satoshi scene scenes sequences shape shows sign signposts signs simulated single smart society somewhat spatial spectral spie springer square sructural state stop study subset supervision symp symposium system systems tech technique temperature that there they this three thus time total tracking traffic transformations triangular typical under using vantageously variations vehicle vehicles verlag vision visionbased were what which will with without work zadeh
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990557	102	Relief Mosaics by Joint View Triangulation	acha acknoledgements acknowledge acquiring alignment also anandan animation applications approach ayer based been bill bishop bodelet camera captured chen clusters cohen computer concentric conference constrained construction creating currently cvpr demonimages depth discussions dominant drettakis duanhuang dyer edge efficient environment environments estimation eurographics example extending field finci finkelstein foundation fruitful full general global gorkani gortler greatly grzeszczuk hand hanrahan held help herman hughes iccv ieee image images impostor interpolation irani joint large layered levoy lhuillier light like local lumigraph maciel manifold manipulation many maps mcmillan model modeling more morphing mosaic mosaicing mosaics motion multiperspective museum navigation only panoramas panoramic paper pattern peleg pipe plenoptic proceedings produce projection quan quicktime real recognition references refinement rendering representation representations rousso salesin sawhney scnenery seitz sequences shade shirley shum siggraph sillion society strated synthesis system szeliski textured thank thayer their this time triangulation triggs universal urban used using video view virtual vision visual visualization williams with wood would
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990520	68	A Nonparametric Statistical Comparison of Principal Component and Linear Discriminant Subspaces for Face Recognition	acknowledgements advanced agency algebra algorithm algorithms altered approach approximations avoiding been belhumeur between boost boult carlo choices class common compared computer conclusions conference confidence contract created cvpr dabt defense differences different distance distributions each eigenfaces establishing estimates evaluations explain explore face feret feretfaceimages fisherfaces fully further future gallery given have here hespanha ieee imagery images important intervals isbn kriegman likely linear linkages method metrics micheals monte more normal over pattern performs probability probe procedures proceedings process projec projects prove rate rates recognition references relative reliance repeated required research samples sampling several similar society somewhat specific strength subspace subspaces summary supported surprisingly synthetically tested tests than that their theirs this trained training under uniformly using variance variant vision while will work worse zhao
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990476	26	Learning Flexible Sprites in Video Layers	able academic adelson algorithm amounts analysis anandan appear applications approach available bayesian believe black click clustering computation computer conference cvpr data deform deformation dempster derived detection dimensionality discontinuities during efficient expectation extraction fast fleet flexible flow fourier frame frey from ghahramani graphical ground hinton http ieee image images incomplete incremental inference integrated intelligence international introduced introduction invariant jaakkola jepson jojic jordan journal june justifies kluwer kumar laird layer layers learned learning lighting likelihood machine maximization maximum methods mixture model models motion moving neal noise norwell object once optical other pattern point powerful probabilistic proceedings process processing prove publishers recognition reduction references removal representation representing royal rubin saul sawhney second sensor sequences significant society sparse sprite sprites stabilization statistical summarization szeliski that through tool torr tracking transactions transformation transforms used utoronto variants variational video view vision wang which will with
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990655	144	On Representing Edge Structure for Model Matching T.F. Cootes, C.J. Taylor Dept. Imaging Science and Biomedical Engineering University of Manchester, Manchester M13 9PT U.K. t.cootes@man.ac.uk	about academic accuracy accurate achieved active analysis ance another appearance appeareuropean appears applications approach areas audio automatic based because belief berlin between biometric bmva bosch both boudewijn bristol british built burkhardt cardiac case cessing changes cognitive colchester compared computer conclusions conditions conf conference continuous contour cootes currently cvpr data database deformable degradation descriptions detection difference different directional discussion distinctive echocardiograms edge edges editors edwards eigen eigenfaces encodes endocardial equal error errors experiment exploring expression expressions extended external face faces factors from further geest gesture give giving glaseby gradient have highlight hond however ieee image images imaging improve improved improvement improvements individuals intelligence intensities intensity international interpreting investigating isbn japan journal june kamp kittler ldabased lead learning leieveldt lelievedt lighting likelihood linearly london luettin machine maitre many mardia matas matching measurement medical messer mirmehdi mitchell model models moghaddam more motion neumann neuroscience neutral nijland noisy normalised noticable novel object orientation pages pairs particularly patch pattern penalized pentland performance performed personal press probabilistic proc proceedings propages proposed rate recognition references reiber reliable represent representation representations responses results rigid robust robustness royal scores search segmentation sept sequences sets shape showed shown significant significantly similar slightly society sonka spacek spie springer statistical statistics strong structure such suggest suppress suspected table taken taylor tends test testing texture than that there third this thomas time training transactions turk type under used useful using variability variation verification verlag video vision vison visual volume vtsdb warping were when where which will with worse
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990453	4	Bayesian Learning of Sparse Classifiers	aaai albert algorithm also amer anal analysis annals applications approach approximation artif assoc atomic barber based basis bayes bayesian berger between beyond binary bishop breiman cambridge chapman chen cherkassky chib classification classifiers comp competition comput computation computer concepts conf conference correspondence cristianini current cvpr data decision decomposition dempster discovery discriminative distributions donoho dordrecht duin editor editors empirical energy entropy estimation fahrmeir figueiredo focus friedman from gaussian generalized girosi graph graphical hall hastie heidbreder ieee image impractical incomplete independent inference info informative intell intend interest isbn issue jain jeffreys john jour kaufmann kernel kimeldorf kluwer knowledge laird lange laplace large lasso learning leen lewicki likelihood linear ller london mach machine machines mackay math maximum mccullagh meth method methods mining model modelling models morgan mulier multivariate neal nelder networks neural nips noninformative normal nowak nystrom observational olshen other overcomplete oxford patt pattern philadelphia poggio polychotomous prediction press prior priors proc proceedings processes pruning publishers pursuit recognition references regression regularization relevance representations researchers response review ripley robust royal rubin rubinstein saunders scient seeger sejnowski selection sets shawe shrinkage siam sinsheimer smoothing society solla speedup spline splines springer stat statistical stochastic stone support systems taylor their theory third this tibshirani tipping trans trees tutz uncert university using vapnik variational vector verlag versus very vision wahba wavelet wiley williams with york
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990656	145	Pairwise Coupling for Machine Recognition of Hand-Printed Japanese Characters	about accuracy acknowledgments additionally advances advantage advantageous advantages algorithms also anal analysis another application approach assignment attaining attains available based bayesian bengio besides best biased block braun buhmann building burges call cambridge character chinese class classes classical classification classifier clear code collobert compared computational computer conceptual conclude conclusion conference costs couple coupling cvpr dealing decision demonstrated discriminant discrimination discussions douglas duda effectively entropy estimate estimation even excellent experiments explicite fisher fixed flannery focused forward framework frieauff friedman fruitful functions general geoscience given handling handwritten hard hart hastie have hermes highly hilbert hiragana hoerl idea idiap ieee imagery implementation imposes information intell interior international interpretation into introduction isbn jaakkola jebara joachims jordan kearns kennard kernel kernelized kernels kimura kopf label labels land landsat large larsen learning leen letters level linear linearly list ller loqo lower mach machines main making march martigny maximum meila membership method methods michael mika miyake modified moreover multi multiclass networks neural nkda nonlinear nonorthogonal number numerical only optimization outperformed over packages pairwise paper particular patt pattern patterns point polychotomous possible posterior practical predicted prediction presented press probabilistic probability problem problems proc procedure proceedings processing programming provided puzicha quadratic quantify recipes recognition reduce references regression remote ridge robert ronan roth rules samy sara scale scene schol sensing signal significantly single smola society software solla sollich space stanford steinhage straight subspace suited superior support switzerland symposium systems takashina tech technique technometrics teukolsky thank that this thus tibshirani training trans transactions trevor tsuda tsuruoka uller uncertainty university usage user using vanderbei variant vector vetterling vision ways weighted weston which wiley wilson wish with within
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990638	129	Single View Modeling of Free-Form Scenes Guillaume Dugas-Phocion Jean-Sebastien Samson Steven M. SeitzÝ	adaptive addison adelson analysis animation anjyo applications approximately arai architecture based basis blanz bovik burt cambridge chen code cohen columns comm compact company compaq comparison computer conf conference constraints construction corporation criminisi cvpr debevec dept depth different discontinuites dorsey durand editing engineering estimation examples faces fast feiner figure flannery focus foley from functions gortler gradient grail graphics grid grids height hierarchical horn horry http hughes ieee image images interace interactive interpolation into inverse involving isbn iteration iterations kang koenderink laplacian left local make malik math mesh methods metrology model modeling models moments morphable mosaics motion multigrid nakagawa nayar novel number numerical original paint painting pami panoramic pattern performance phil photo photographs phys pictorial picture points practice press princeton principles problems proceedings project projects publishing pyramid quadtree recipes recognition references regularization reid relaxation relief rendering research right scenes science sciences shading shape show shum siggraph single society solving specified spectral spidery splines super surface symposium synthesis szeliski taylor tech terms terzopoulos teukolsky texture textured tour traditional trans transformation univ university user using variational vetter vetterling view vision visual washington wavelets website wesley williams wireframe with zisserman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990549	95	Efficient Computation of Adaptive Threshold Surfaces for Image Binarization Ilya Blayvas, Alfred Bruckstein and Ron Kimmel Computer Science Department, Technion Institute of Technology Haifa, ISRAEL, 32000 {blayvas,freddy,ron}@cs.technion.ac.il	academic adelson algebraic analysis approach automatic binarizaiton binarization biomed boundary briggs bruckstein burt chow cineangiograms code communications compact comput computation computer computing conference courant crowley cvpr detection difference document evaluation fast figure from graphics hilbert ieee image images intelligence interscience isbn journal kaneko kimmel laplacian left machine mallat mathematical method methods multigrid oxford pami pass pattern philadelphia physics press proceedings processing publishers pyramid recognition references relaxation scientific segmentation siam signal society southwell stephane stern submitted taxt theoretical tour transactions transform trees trier tutorial university ventricle vision wavelet with yanowitz yavneh
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990573	110	A tool for decomposing 3D discrete objects	acceptable algorithm analysis anyway applications approaches arcelli aspects axis baja based beginning belonging biederman body borgefors bottom braccini case cognition complex components computational computationally computer computing conference connected consisting constitute constrained convenient cost cvgip cviu cvpr cylinders decomposed decomposing decomposition decompositions description desirable different digital dimensional dimensions discs distance dorst easy effective elsevier euclidean even example examples figure finally finger fingers forefinger form from future gives good graytone great hand have hoffman human ieee illustrated image implementation included individual interest into introduction isbn kernel kimia kong lack left likely mentioned method middle more nackman number nystrom object objects obtained obtaining other overlapping pages pami paper part parts pattern perceived performance performs pizer possibility precision proceedings processing properties pseudo psychological real recognition recursive references representation respect result results reverse review richards right rosenfeld same sanniti satisfactory scenes seconds seems serino several shape shown siddiqi signal simple sized skeleton society space springer such survey symmetric synthesising tested theories theory thiel this though three thumb tool topology touching trans transform transformation transforms understanding unique using verbeek verlag visible vision visual voxel weighted well which with young
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990653	142	Affine arithmetic based estimation of cue distributions in deformable model tracking Siome Goldenstein Christian Vogler Dimitris Metaxas V.A.S.T. Laboratory - University of Pennsylvania 200 S ¿¿rd Street, Philadelphia, PA 19104 USA siome,cvogler,dnm @graphics.cis.upenn.edu	abstracts academic acknowledgments adaptive addition addressed affine afosr aided albedo algebraic algorithm algorithms allowing allows analysis andrade annual another anthropometric application applications approach approximated arithmetic august author authors available award azar based because between biomedical blake bound boundaries bounds branch brasileiro brazilian breast broida cameras campinas career carlo centered chellappa cientfico cnpq coloqui comba combination combining come compression computer computing conclusions conditional conference confidence conselho constraints control correlation covariance cover cues curve cvpr data decarlo deformable deformations dekker density depicts depth desenvolvimento design different discarded distribution does domain during dynamic eccv elegant engineering entire estimate estimation even example face fast feller figueiredo figure filter filters first fitting flow form forms framework from functions further future gamm global graphics hall hand hansen have heart high hill iccv icondensation ieee ijcv imacs image images imaging impa important improve inference information initial institute integrating integration interaction international intersections interval interventional introduced introduction introductory isard isbn iwaarden john journal july june kalman kamberova kashiwagi kluwer knowledge large learning leclerc level like local location long loss maekawa mahfoudi mandelbaum matemat maybeck mcgraw mechanical medical meeting messine metaxas method methods minimax mintz model modeling models moore more motion multi multidimensional nacional noisy nolta nonlinear number numerical numerics object obtain ones open operations optical optimization other over pages pami paper papoulis parameter parameters park part patrikalakis pattern physics planning plate predicting prediction predictive prentice press prior probabilities probability problems proc procedures proceedings process processes propagate propagation publishers random recognition reconstruction references regions report required restricted results robust rochester rules samaras scales scaling schnall science scientific self september shading shape sherbrooke short should siam siggraph small snyder society solution solve some sons sources spanned statistical stereo still stochastic stolfi stone supported surface symposium technical techniques tecnolgico term thank that theory there they this tracking tradeoff trucco unconstrained under understand unifying univ using validated variable variables variational verri vision visual volume wall well will willey with within without work would young zero
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990498	46	Extraction of illusory linear clues in perspectively skewed documents Maurizio Pilu Hewlett-Packard Laboratories Bristol, BS34 8QZ, UK maurizio pilu@hp.com	able abundant algorithm aligned along although amongst analysis analytic application approach artifical artificial automatic background baird based bayesian bolles boundary bruce bundle cambridge camera cameras candidates capturing choi clark clues common compass components computer computing conclusions conf conference correct correctness coughlan cvpr cylinders dance data detected detecting determine determining direction distorted document documents edges edition electronic endeavoured establishing estimation examples fact figure finding fischler fitting focal fojinawa followed found four from fujisawa full general geometry given gorman green hand haralick hartley hashizume have held hence higashino hong horizontal however icpr identification ieee illusory image images imaging impossible indeed inference intelligence internation international into inverse isbn january joint journal just know knowledge known kwon laypattern lenght length letters likely lines location machine makes manhattan margins messelodi method mirmhedi model modena monocular more morphological much multiple nakano newman normalization novel november obtaining ones only order orientation oriented originates other output page pages paper paragraph pattern perception perform perpendicular perspective perspectively point practice presents press printing probability proceedings produce projection projective range ransac rare real recognition recovery rectification rectify references relations reliably retrieval rosenfeld scene second section segmentation separate september several shima since single situations skew society spectrum spie storage structure such summary surfaces taylor test text textures that then theory there these they this three through undistorted university upright used using vanishing vertical view vision visual volume what when whether with world yuille zappala zisserman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990651	140	Matching of Double-Sided Document Images to Remove Interference Qian Wang, Chew Lim Tan School of Computing, National University of Singapore, Kent Ridge, Singapore 117543 citwangq@nus. edwsg, tancl@comp. nus.edu.sg	accuracy analysis application approach april astronomical attribute automatic back background based binarization binarized california canada cancellation canny chang character chee communication components computational computer conference cvgip cvpr cybernetics dengel detection document documents double duplex edge entire evaluation extraction features figure final from front gaurav gray haseand histograms hoch ieee image imagebinarization images india interference interfering international isbn jager junker kato level matching method murtagh negishi noise noisy original otsu pacijic page pair pami pattern point precision printed proc proceedings processing publications recall recognition referencesystem removal results scans selection sharma shen show sided signal society south srihari strokes survey symposium system techniques texture threshold thresholding through trans vision wacv wang watanabe workshop
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990497	45	Navier-Stokes, Fluid Dynamics, and Image and Video Inpainting	akad album amsterdam anal analysis appl applied arch award awards ballester based beghdadi bergen berlin bertalmio bertozzi besserer books boukir braverman breach browning buisson cambridge career caselles catte cbms chan characterization chicago coefficient coll collins colorado combining commissar commun comp company computational computer conf conference corfu cvpr dauphine depth desocclusion diperna disocclusion domain early easel edition efros emile engineers ensembles equations fast fatemi filtering filtrage fitzgerald flow flows fluid fort frequency galdi gordon grants graphics greece greengard haar handbook heeger helmholtz henry hirani holland holt ieee image images incompressible information intelligent interactive international introduction isbn joint journal joyeux judovich july june kato kelvin king kokaram kreiss ladyzhenskaya learning leung level lines lions london lord mach magnitudes majda male march martin masnou math mathematical mathematics mech methode methods monatsber monteil morel morris motion mumford natural nature navier nitzberg niveau noise nonlinear north nostrand osher painting parabolic paris patt pattern pecase peyret philadelphia philosophy photoshop phys portilla presidential press problems proc proceedings processing program publishers publishing pyramid ravished rayner recognition references reinhold removal renardy report research restorer retouching ringach romeny rudin sapiro sbert science scientists second segmentation september series sethian shapley shen shiota siam siggraph simoncelli society spatial springer stanford statistics steady stokes synthesis systems tang taylor temam texture theory these thomann totsuka tracts trans ucla univ vandyke vanishes verdera verlag viergever viscoelastic viscous vision vorticity walden wavelet weickert wolibner worldwide york zeit
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990658	147	Bayesian Color Constancy for Outdoor Object Recognition Yanghai Tsin Robert T. Collins Visvanathan Ramesh Takeo KanadeÝ	amer analysis approach ballard barnard basic bayesian between brainard buluswar calibration characteristic chips cohen color colors colour comput computational computer comupter cone conference constancy correlated curves cvpr cybernetics daylight dependency descriptions diagonal dimensionality distribution divergence draper drew engineering enough european evaluation finlayson forsyth foveal freeman function functions funt good hall hallikainen healey hebert hill ieee illumination images imaging inidexing intelligence international isbn iverson jaaskelainen jansen january jersey journal judd july june kanade learning lights linear macadam machine maloney martin mcgraw meer mitchell model models multiple munsell normalized novel november numbers october ohta outdoor pages parameters parkkinen pattern photopigments pokorny prentice proceedings process psychon ramesh recognition recovery references reflectance resampling robust rosenberg sampling scenes segmenting sensitivity slater small smith society solar spectra spectral stage statistical suffice surface surfaces swain synthesis system techniques technology temperature theory thrun trans transaction transforms tsin tsukada tutorial typical using vision wandell what with wyszecki york zmura
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990537	84	Bernd Heisele	acknowledgements additional analysis authors based beymer biological blanz brunelli burl cambridge center classifiers cluttered compaq component computational computer conference contract cvpr darpa detection eastman face faces fast feature features finding foundation graph gray hawaii heisele hierarchy honda ieee images intelligence isbn kodak labeled learning leung like machine matching memo models mukherjee national object partially pattern perona poggio pontil pose proc proceedings provided providing random recognition reduction references research scenes science serre society sponsored still support templates thank transactions under using varying versus vetter video vision would
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990513	61	A Confidence Measure for Boundary Detection and Object Selection Eric N. Mortensen1,2 (enm@cs.orst.edu) William A. Barrett1 (barrett@cs.byu.edu)	algorithm algorithms anal approach assessing baker barrett based beucher binford blur bowyer cambridge canny color columbia compass composition computational computer conference connexion contour contours contrast control cvpr design detecting detection detectors digital digitized dijkstra edge edgedetection edges efficient elder enhancement estimation evaluation fairfield feature four geodesic gleicher gmip graphics graphs heath hierarchical hildreth hueckel icpr ieee image immersion intelligent interactive isbn jour journal july june lantu local locates london marr mathematik method model mortensen motion najman nalwa note numerical numerische operator overview pami parameter pattern performance pictures press problems proc proceedings processing real recipes recog recognition references relative robust royal ruzon saliency sanocki sarkar scale schmitt scissors segmentation sept simulations snapping society soille spaces tabbone techniques theory thesis time toboggan tomasi university vincent vision visual watershed watersheds which with workshop ziou zucker
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990458	9	Dense Image Matching with Global and Local Statistical Criteria: A Variational Approach Gerardo Hermosillo Olivier Faugeras INRIA Sophia Antipolis 2004 route des Lucioles BP 93 Sophia Antipolis Cedex, France Gerardo.Hermosillo,Olivier.Faugeras @inria.fr	about acknowledgements action after algorithm algorithms alignement also alvarez american analogy analyse analysis anandan appear appeared application applications applied approach arcs artificial asian assisted association aubert automatic ayache based before bergen bland boes bottom brain breast brezis british cambridge canada center chefd christensen ciarlet cibernet ciencias colchester columbia computer computerassisted computing conacyt conference constraint control corfu correlation correspondences cuadernos cvpr deformable deformation delp demons dense density deriche diffeomorphisms differential diffusion direct discontinuities displacement displacements distributions dohi driven dublin editor editors eight elastic elasticity enkelmann equations estimation european evaluation evans face fast faugeras fields fiels figure first flow fonctionnelle framework from function gaens general global goos graduate grant greece grimson group groups guided guimond hanna hartmanis hata hawkes hayes hermosillo hierarchical hingorani holland homeomorphism hotel human icas ieee ijcv image images imaging information inria instituto intelligence intensity interesting international intervention intra investigation irani ireland irmf isbn issue italy iterative joint jolesz journal july june kanade kikinis kinematics kornprobst large leach lecture left level leventon lucas machine magnetic maintz malandain mapawamo margherita masson matching math mathematical mathematics maximization maxwell medical meijering methods metric meunier meyer miccai miller modal modality model motion multi multimodal mutual nagel nchez nonrigid north notes number operative optical pages partial partially parzen pattern pennec point points posedness practice preserving probability problems proceedings process processing published rabbitt ratio recognition reference references registration regristration reliable report resonance respective revised right rigid roche ruck sandini santa science selection september sequences similarity singapore smoothness society some special spie springer statist statistical stereo studholme studies suetens summers supported surgery symposium target technical technique tecnolog template textbook their theori theory thirion this transactions trouve universitario university using vancouver vandermeulen vannier variational vector verlag version viergever viola vision volume warfield warped warping weickert well wells while with work workshop younes
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990632	124	Generalized Dynamic Programming Approaches for Object Detection: Detecting Spine Boundaries and Vertebra Endplates Guo-Qing Wei, JianZhong Qian and Helmuth Schramm Imaging Department,Siemens Corporate Research, Inc. 755 College Road East, Princeton, NJ 08650, USA Medical Solutions, Siemens AG, Erlangen, Germany	about above accuracy achieve active algorithm also although amini analysis angular applications applied applying approach area artif assisted automatically baker balloons barrett base based biased binford blurred boundaries boundary cohen comparing computer computing conf conference consisting constraint constraints context contour contours cootes copper correct cortex costa coupled crosses curve curves cvgip cvpr data deformable deformities degrees depth detected detecting detection disturbances dual dubisson ducan dynamic edge eigenfaces endplate endplates errors especially example face falcao figs figure find finds fitting found four from geiger global graham graphical gupta height help helps hingorani human iccv ieee ijcai image images imaging improve improves ineractive information intell intelligent intensity inter intercurve interpolate intervention intra isbn jain jolly kass lane left likelihood line lines live local localization maggs manually marked matching maximum measuremnt medical method miccai middle missing models mortensen motion next noisy nonrigid note ones only orientation pami parabolic paradigms part path patient pattern pentium pentland pixels points position positions presence problems proc proceedings processing programming propagation qian quantitative real recognition recovering references reflects results right samarasekea schramm schultz scissors seconds seen segment segmentation selected serious shall shape sharma shortest shown shows small snakes society solving some spine spurious staib steered stereo straight strong surface tangent taylor templates terzopoulos tested that their this threshold took tracking training trans true turk udupa unable understanding used useful user using variational verdon vertebra vision vlontzos well weymouth when wire with without witkin zeng
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990539	86	A Radical Approach to Handwritten Chinese Character Recognition Using Active Handwriting Models D. Shi S. R. Gunn R. I. Damper Image, Speech and Intelligent Systems (ISIS) Research Group, Department of Electronics and Computer Science, University of Southampton, Southampton SO17 1BJ, United Kingdom Email: [ds99r,srg,rid]@ecs.soton.ac.uk	active algorithm algorithms analysis application applied artificial ashida attention based borgefors chamfer chang chansarkar character characters chinese chung classification component computation computer conference contour cooper cootes cvpr cybernetics damper descent designing document dong dynamic edge evolution extraction fukushima garaham generation genetic germany global gradient gunn handwriting handwritten hierarchical huang humans hutchinson hybridization ieee ijcnn imagawa image intelligence interactive international invariant isbn joint jolliffe journal kass liao london machine matching mathematics method methods modeling models multifeature multilevel neocognitron networks neural offline optical optimization parameter parametric part pattern principal proceedings radical recognition references retrieval roychodhury sampson seattle selective shape shyu singh sixth snakes society springer submitted system systems tang taylor terzopoulos their third towards training transactions transformation tuning tunneling understanding verlag vision wang with witkin writing yeung york
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990625	120	Epipolar Geometry Estimation for Non-Static Scenes by 4D Tensor Voting	accepted adam affine algebraic algorithm amsderstam analysis anandan application avidan background ball barbara baseline blue bombay camera clarendon coded colored combine component components computational compute computer computethealignmentmatrix computing conference corresponding count cvpr davis defense detection determining dimensional direction discarded distributed eccv eigenvectormatrix elseviers england epipolar estimate estimation excepttheisotropicball extraction feature fgote figure first framework from fundamental gennormalv gennormalvote geometry global green hartley higgins iccv ieee image images independent india indicated integrating intelligence international into isbn january joint journal june kneebone last left lines local longuet machine matches matching matrix medioni methods mosaics motion movement moving murray nature noisy number objects outliers outtensor oxford pages pair pami pattern plate point press pritchett proceedings projections projective random randomuniformpt recognition reconstructing references rejection results return review right rivlin robust rotations rows samples santa scene scenes science segmentation semple shimshoni show society space sphere static statistical stereo stick sufficient tang tensor their then torr transactions transformvoter umbrella uncertainty uniformly unit vecvote vision votee voter votersaliency voting while wide with zhang zisserman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990511	59	Model-BasedCurveEvolutionTechniqueforImageSegmentation	active activecontourswithoutedges additiontoshowingthe algorithmsbasedonhamilton alsofrombwh alsoshows analysisandmach anda andj ando andpatientb andratedthecomputer anintegrated anotherradiologist astatisticalapproach betterthanthecomputer bottomframe boundaryndingwithcorresponvisionandpatt boundaryndingwithparametri callydeformablecontourmodels chakraborty chanandl comm comp comparedtothesegmentationbyaninteractive computer conclusions conference convergestoalocalminimumneartheboundariesof cooper cootes cvpr dcardiacmrisegmentationalgorithm denceusingstatisticalshapemodels dependentspeed dhandsegmentations dimagesofeach dmodelofthesegmentedprostateisshownontheright dprostategland dshapemodelsoftheprostateglandobtainedbasedonstacked dslicesofthemrdataset duncan eight faugeras figure frontspropagationwithcurvature graham greencurves grimson gureshowthesegmentationbyouralgorithmand ieee ieeeconf ieeetrans intell isbn iscompared jacobiforj lconf leftventriclesegmentationresultsofcardiacmri leventon math mumfordandj neitherofwhicharepart ofcomput ofthe ofthecurve ofthetrainingdatabase onimageprocess optimalapproximationsby osherandj outwardtocapturemoreandmoreofthelwvoalued patientrepresentconsecutiveaxialslicesoftheprostate patt pattern phys piecewisesmoothfunctionsandassociatedvariationalproblems principalmode proceedings prostatesegmentationofpatienta pureappl radiologist ratedtherst recog recognition redcurves references regioninthetransformeddatawhichcorrespondsto segmentationofpatientbtobeslightlybetterthan sethian shah shapemodels shapevariabilityoftheprostate shospital showstheprostatesegmentationresultsoftwopatients society ssegmentationofpatientatobeslightlymulations staib staibandj statistical tationalgorithmbasedonanimplicitrepresentation taylor thegreencurvesshowthesegmentationbyaradiologistfrombrighamandwomen theirtrainingandapplication themeanshape theprostatecorrespondingtohighvaluesinthetransformeddata theradiologist theredcurvesinthe thesegmentationbyouralgorithm thesegmentationbytheradiologist thesegmentationoftheprostateintwelvecontiguousshapeinuenceingeodesicactivecontours thisgure thisrepresentationissetinaneuletosnakesforbimodalandtrimodalimagery toappear topframe tothesegmentationbyouralgorithm trainingdata tsai using variance varianceofthe vese vision visionandimageunderstanding visionandpatt wangandl wehaveoutlinedarobustandecientsegmen willsky workshoponbiomedicalimageanalysis yezzi
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990650	139	Automatic Estimation of the Projected Light Source Direction Peter Nillius Jan-Olof Eklundh Computational Vision & Active Perception Laboratory (CVAP) Department of Numerical Analysis and Computing Science Royal Institute of Technology (KTH), S-100 44 Stockholm, Sweden	able accidentally achieved adaptive after algorithm also ambiguity angle angles application approach april automatic background bayesian been belhumeur better boundary brightness calculated candidate case causes challenging circle classified classify classifying color combined compared computation computer conference considered continuation contour contours cope correct correctly could cowell crucial cvpr dawid default determination dielectric diffuse direction distribution distributions done dotted edges errors estimate estimated estimates estimating estimation example examples experiments expert fail fails figure finding first from fully fusing future generalization give gives good grouping happen heuristic hill horn iccv ieee ijcv ikeuchi illuminant illumination image images implications improvements incorporate incorrectly inference inlined introduction isbn january jensen josa kanade klinker kriegman lamber lauritzen learning light limited lines machine main mcgraw message model modeling more moreover multiple nayar network networks nonoccluding note november occluding often oren pattern pentland physical picked probabilistic probabilities probability proc proceedings produce promising propagation properties real reasonable recognition references reflect reflectance reflects regions relatively relief respects results robot robustness same sato scenes second shading shadow shadows shafer shape should show shows single smooth society some sort source sources spiegelhalter springer stage still suitable surfaces systems test tested than that theory thereby this tian true understanding unknown used using usually values variance vega vision when which whole will winning with wolff works yang yuille zhang
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990461	12	A New 3-D Pattern Recognition Technique With Application to Computer Aided Colonoscopy	abdom academic acar acknowledgement addition aided algorithm also analysis application applied apply approach appropriate area areas automated axial barish based beaulieu bioengineering biomedicine biosciences blinded boston burak campbell cancer chris citeseer clarke classes classification clin colin colography colon colonography colonoscopy colorectal comparison compression computer conduct conservative conventional current data david debatin detection detections detector directions discussions display distinguished duda endosc endoscopic engl evaluation experiments false feasibility fenlon ferrucci free from future gastroenterol gastrointest gastrointestinal generous gersho glazer gokt grant graph gray hart high html imaging information inspiring international invariants investigation jeffrey john journal judy karadi kernel kluwer laufer learning like luboldt lung malley many medical method methods modes moment more munich napel nature need nodule nunes objects oldenbourg other output page paik panoramic part pattern performance philadelphia polyp polyps positive possible press program pusanik quantization radiology ratio recognition reduce references report response ross sandy saunders scene scholk schroy seed sensitivity signal similar sons springer stanford statistical statistics status steinauergebauer studies study such summers support supported survey symposium techniques textbook thank theory there this thoeni thus tomasi transactions university vapnik vector verlag vining virtual visual volumerendered where wiley wingo with work would york
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990577	114	A Linear Iterative Method for Auto-Calibration using the DAC Equation 	absolute adjustment after agapito algorithm algorithms analysis applicatio applications armstrong aspect assumption astrom austria auto autocalibration barcellona based before berthilsson between bougnoux bundle buxton calibation calibration cambridge camera cases change cipolla collineations complex computer computing conf conference constraint constraints critical criticism csurka cvpr derived distribution dublin eccv editor editors eigendecomposition environments epoch equations eration error euclidean european experiment experiments factorization faugeras figure fitzgibbon fixed flexible focal from fundamental gool greece hartley hayman heyden hong horaud iapr ieee image images india infinity inria intelligence internal international ireland isbn iteration iterative japan kahl kerkyra koch kruppa large lecture length linear locations luong machine malis matrix maximum maybank mclauchlan mendonc method metric minimal modern modulus monocular motion motions multi multiple mumbai notes oosterlinck orthogonality pages parameters pattern pixel planar point points pollefeys ponce practical practice press principal proc proceedings projection projective quadric ratio real recognition reconstruction reconstructions references reid report reprojection research retangular revisited sandini scale scene scenes science search self selfcalibration sequences simple simultaneous singular situation skew smile society space spain sparr spite springer springerverlag stereo structure structures sturm svals synthesis technique theory through tokyo trans triggs triplets uncalibrated under unknown upgrades using values variable varying verlag video vienna view vision volume with workshop xtaylor zeller zero zisserman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990623	119	A New Analysis Framework for Relevance Feedback-Driven Similarity Measure Refinement in Content-Based Image Retrieval	aligns architecture artificial based bhanu cambridge case cbaivl cbir closely cohen color combining common computer conf conference content contentbased correlograms cristianini cvpr data databases dissertation driven empirical examples feature feedback framework function goals huang ieee image information intelligence introduction isbn ishikawa jones learning limited local machines maps method methods metrics mindreader more multimedia multiple objective operation other outperforming particularly pattern persistent presented press proc proceedings proposed querying real recognition references relevance results retrieval saliency shawe show similarity society suited supervised support system systems taylor through training true univ university vanderbilt vector vision vldb well which with world
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990559	103	Real time 3D template matching Frédéric Jurie and Michel Dhome LASMEA - CNRS UMR 6602 - Université Blaise Pascal - 63177 Aubière cedex - France	active actual after analysis angle appearance approach approximation april around article articulated athitsos axis based been belhumeur beyond black brunelli canada cascia case code computer conference cootes cube cvpr darrell davis decomposition dementhon dhome difference diverges edwards efficient eigentracking erence error essa esti european experiment experimentally fast figure filter former free freiburg fronto geometry germany gesture gleicher hager have head ieee illumination image intelligence international interpolated introduced isbn january jepson journal july june jurie limitation lines localization machine made mapped matched matching mated memo mentioned method model models more motions object objects observed october only pages pami parallels parametric pattern perturbation pose position previously proc proceedings processed projective proposed real recognition references region registration relation relatively reliable representation robust rotated rotations sclaroff second sequence showing shown shows small society spatial specific straightforwardly suited task taylor template textured than that this time tolerate tracking trans translations under used using vancouver variations varying vertical very video view views vision vison which while with
http://csdl.computer.org/comp/proceedings/cvpr/2001/1272/01/127210998abs.htm	156	3D Head Tracking UsingMotion Adaptive Texture-Mapping Lisa M. Brown IBM T.J Watson Research Center Hawthorne, NY 10532	acoustics akamatsu akarun alpaydin analysis application approach automatic based behumeur bibliography broadcast chellapa chung component computer conf conference corfu cvpr cybern dark demir detector dotted face faces feature feng figure filter france from frontal gabor generative georghiages gesture gestures greece grenoble ground head human humans ieee illumination image information instanbul integrated invariant isbn june kouzani kriegman light line linear malsburg many march model models okada pattern pcmap pose principal proc proceedings processing real recognition recognizing represents responses results robust sammut sciences senior sept shoulder signal society solid speech stage synthesis syst system systems time towards tracker tracking trans truth turkey under using variable variations video view virtual vision with workshop yuen zhao
http://csdl.computer.org/comp/proceedings/cvpr/2001/1272/01/127210689abs.htm	154	Precise Omnidirectional Camera Calibration Dennis Strelow, Jeffrey Mishler, David Koes, and Sanjiv Singh Carnegie Mellon University dstrelow, jmishler, dkoes, ssingh @cs.cmu.edu	analysis angular applied bombay calibration cambridge camera cameras carnegie carolina catadioptric chahl communication computer computing conference corporation cvpr daniilidis design epipolar equi european extending fifth flannery freiburg from geometry germany geyer gluckman hawaii head herman hilton hlavec http ieee image imaging institute intel intelligent international isbn island journal june kang least library march mellon mishler motion nayar noncentral nonlinear numerical october ollis omnidirectional open opencv optics pages pajdla panoramic pattern pittsburgh pixel press proceedings recipes recognition recovering references reflective report representation research robotics robots scientific self september seventh shape singh society source south squares srinivasan stereo streams strelow surfaces svoboda systems szeliski technical teukolsky university using vetterling vision visual wailea
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990525	73	A Spherical Eye from Multiple Cameras (Makes Better Models of the World)	absolute accurate advantages algorithm algorithms allows aloimonos also ambiguities argus based behaved better between beyond brings brodsky built calculate calibrated calibration called camera cameras case centers clearly coincident complicates computer conclusions conference consistingof constraint construction cvpr daniilidis data defined definite department depth descriptions developed different directions distance each editor eigen eigenvalue eigenvalues epipolar error established estimate estimation fact fermuller field fifth figure force from full function german germany have here hessian highresolutioncameras ieee imaging individual informatics information international journal karlsruhe large larger largest literature metric minimum motion much nodal noise none note object obtain other over parameters part pattern plenopticfunction point possible proceedings processing projection provides recent recognition recovering recoveryof references remarkably results rich robustness rotation same sampling sensitivity shown significant simpler since sitivityinstructurefrommotion sixth smallest society solution space spetsakis stereo structure support synchronized system techniques than that theoretical thesis they three thus together translation understanding university using value very videos view vision well where which while without work
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990646	136	On the Perceptual Organization of Texture and Shading Flows:	able academic agrees album along amandus american analysis analyzing applied background based bayesian beck biological book bottom cambridge case caselles cavanagh classification coherent company compatibilities computation computer conf conference consistency contour cortex cover create curvature curve curved cuts cvpr cybernetics data dense detection diff diffusion direct direction discontinuity discrete dots double dyke early edge editor editors effect elementary ellis equivalent error essays estimation field figure fills fingerprint flat flow flows fluid foundations framework freeman from further gaustein gedachtnis general geom geometric gestalt given glass gradient graphics hamel harmonic helicoidal hermann hope human hummel ieee image images inference initial intelligence intensity interest isbn iverson jahresber jain kanizsa kass keeble kegan kingdom kluwer labeling laciness lectures locally lowe machine malik manifolds many maps marr mathematical mcgill measurements middle minimal missing models moire morgan motion natural nature neill nitsche noisy normalized nose note organization organize orientation orientational oriented pages pami parabolic parallel parent particularly pattern patterns paul perceeption perception perceptual performance perona photo plane praeger preserved preserving press primary proc proceedings proceeses process processing projected psych publishers random recognition references region reichel relaxation representation research resolution result resultant right role rosenfeld routledge sapiro schwarz segmentation shading shows singularity smoothly society source stevens structure such superimposed suppressing surfaces system table tang tenenbaum texture that theory thesis todd toward trace trans transparancy underlying understanding univ university useful vision visual volume watanabe wertheimer which while with witkin zucker
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990530	78	1	alexey alternative analysis animating april banana banerine belgium bishop bottom catholic coefficients color compact complex component computer conf conference correlation covered cvpr dana dependent direction early engineering entropy esat ethz fast fields figure filters frame from function genetic ginneken gool graphics histograms http iccv ieee image images interactions irradiation isbn joint julesz koenderink lecture left leung leuven levoy malik mapping march maximum mcallister model modeling multiview mumford nayar neyret notes oliveira pairwise pattern perception plates pollefeys portilla proc proceedings processing psychol quantization random real recognition recognizing references relief rendering scenes schumer science siggraph signal simoncelli skin smile society statistics structure surfaces synthesis synthetic tangerine tech textons texture textures threedimensional topright trans treestructured university using vector viewing viewpoint vision visual visualization volumetric wavelet wilson with workshop zales zalesny
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990506	54	Gait Recognition Using Static, Activity-Specific Parameters	ability academic acceptance accepting achieve adelson also analysis analyzed analyzing angle applied approach april area artifacts audio authentication autoid automatic background based because been best between biometric biometrics bobick body bolle boston boyd bulletin canonical carlsson carter case certain characteristic cognitive combination compare compared completeness computer conclusion condition conditions conference confusion considering create cues cunado curve curves cutting cvpr data database davis demonstrated density detecting detection developing deviation different discrimination displacement display dynamic each eccv eigenspace environments evaluate evidence expand expected extremely face fact falls false familiarity figure figures foreshortening frame friends from gait gallery gathering gaussian gesture gives greece halmstad hanley haritaoglu harris harwood have higher highlights horprasert however http huang human iccv ideal identification identity ieec ieee image individual indoor interesting international isbn jain johnson journals june kerkyra kluwer kozlowski large less letters light little lower mcneil meaning measurement method methods mitpress model modeling most motion moving multi murase must networked next nixion nixon niyogi object obtain operating optimal other outdoor over pages pairs pankanti paper paradigm parameters pattern people perception performance person personal point power predict presented previous probe proc proceedings processing procs psychology psychonomic psychophysics publishers quality radiology rate rates reading real receiver recognition recognizing recogntion recovered references rejection reporting representation represents result results robust sakai seattle sensitive september several shadow shape signal simpler society space spawn standard static statistical steps stevenage stride subjects subset subtraction summint sweden system templates temporal test testing that their there these third this threshold time tracking traditional under undertaken using various varying video videre view vince vision visual walk walker walking were what when where with within without work workshop
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990627	121	Tracking the Optic Nerve Head in OCT Video Using Dual Eigenspaces and an Adaptive Vascular Distribution Model	able acceptable actual additional algorithm algorithms analyzing application approach asilomar automated automatic beam because becker belhumeur between bifurcations biomedical biomedicine black blood botha camera caught characterization chatterjee chaudhuri christopher circle class clinical clinicians cogneuro collins computer computers conclusions conference constructing contrast correctly could currently curved cvpr dark darryl daunicht david dearborn detect detection detections determination developed developing deviation diameter different dimensional direct discriminant draw driver easily eigenface eigenfaces elastic engineering estimating exceptionally expected exploratory extraction extremely face faces faint feature features featuring field fields figure fill filters find first fisherfaces five footage fort free from frontal fundus gandhi gaps goldbaum grants gray group have head hence hespanha heterogeneous hierarchical high however human ieee image images imaging improvement induced information instead intelligent intended interpolated isbn itsc january joao june juwei karhunenloeve katz kirby kriegman lack lacking lades lakshmanan landmark large laser letters line linear lingmin live localization located location locations lost machine made many mapping march markings martin matched matching mean measurement meng method missed misses montage moreover mosaic motion motions movement multiple near nearest nelson nerve nervehead nets networks neural nguyen nieuwoudt nitin noteworthy observed obtain obtaining occurred october ocular onerous ophthalmoscope other pami path pattern paul pentland performance peter pixels poor population pose position presence present previously procedure proceedings processing projection prototype purnell range rapid ratio real recognition record references relative relied repeated requirement requires research resolution responds results retina retinal robust roysam runs sample scan scanning scans science seconds seen sensitive september sequences shen show shown shows signals single sirovich smooth society specific spot spots sridhar stan stewart such surety switching synthesis system systems tanenbaum technology that theory they third thirty this thus time tracing trans transactions transportation true truong turk turner undetected unit upon using vascular vast very vessels video vision watta website well were where which while white with works yong zhang
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990533	80	Scene-Consistent Detection of Feature Points in Video Sequences Ariel Tankus Yehezkel Yeshurun School of Computer Science Tel-Aviv University Tel-Aviv 69978, Israel arielt,hezy @post.tau.ac.il	achieves analytically arbitrary argument attention automatic average based bauckhage better bombay breaking camera camouflage chellappa color comp compare comparison completeness computer conclusions conf conference consistent convexity corner criterion crossing cvpr detectingsalientblob detection detectors each equate estimation european evaluation extraction failed feature features figure focus frame frames frantz from germany good gradient greece have highly ieee image images india intensity interest international intl invariant journal junction june kerkyra laganier landmarks last left length less likeimagestruc lindeberg local localization lowe method methods minimal mohr more morphological motion multi notrack object observing only oriented other over parking pattern performs point points presented procedures proceedings prominent recog recognition references rehrmann relates report right rohr ruzon scale scene sceneconsistent schmid seattle selection sequence sequences show significantly size society spaceprimalsketch specific stability stable step stiehl tankus technical textured than that then this three time together tomasi total toys track tracker trackers tracking traffic understanding uresandtheirscaleswithascale video vision visual well window with without would yeshurun zero zheng
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990636	127	Discrimination of Motion Based on Traces in the Space of Probability Functions over Feature Relations Sudeep Sarkar and Isidro Robledo Vegay Computer Science and Engineering Department University of South Florida Tampa, FL 33620 fg@csee.usf.edu sarkar, irobledo	about access activity anal analysis applications approach articulated assumed based belongie between black bobick boyd claim classifying coding collins computation computer conference consider content contexts correspondence correspondences crowley cutler cvpr cyclic darpa database davis december definitive detection done dyer efficient eigentracking embedding european exciting features field fields flow frames from gait gestalt gesture goddard have here histograms human ieee image intel international introduction invariant isbn jain jepson jogging journal kanade kinds kluwer larger laws libraries lipton little local lowe mach make malik markov matching method models motion motionbased motions movement moving multidimensional natural need nelson network neural nonrigid november object objects olshausen optical page pages parameterized pattern people performance periodic persons pixel polana presented proceedings publishers random real realize receptive recognition recognizing references representation results rigid robust running scale schiele section seitz selinger sequence sequences shah shape single society special state statement statistics surveillance systems temporal testing texture that their time trace tracking trans using video videre view vision walking wilson with without wixson workshop yacoob
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990647	137	Learning Models for Object Recognition Pedro F. Felzenszwalb Artificial Intelligence Laboratory Massachusetts Institute of Technology Cambridge, MA 02139 pff@ai.mit.edu	algorithm applied barrow blumer bolles borgefors cars chamfer class classification compaq comparing computational computer concepts conference correspondence cvpr detection dimension distance distances edge efficient ehrenfeucht faces february formulation gdalyahu hausdorff haussler hierarchical huttenlocher ieee ijcai ijcv image images introduction isbn jacobs jones june kanade kearns klanderman learnability learnable learning lncs matching method minsky nonlearnable nonmetric november object olson pages pami papageorgiou papert parametric pattern perceptrons poggio press probabilistic proceedings real recognition references report representation retrieval robust rucklidge schneiderman september shvaytser society springer statistical system technical techniques tenenbaum theory time trainable using vapnikchervonenkis vazirani verlag viola vision visual warmuth weinshall with wolf
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990501	49	3D Simultaneous Localisation and Map-Building Using Active Vision for a Robot Moving on Undulating Terrain	active after against also angle anyone appear application applications approach areas array asian automatic available based been blake building burgard cambridge camera carried causally chiuso chong close closed computationally computer conclusions concurrent conference consider could course cvpr davison deployments described dissanayake domain dublin durrant editors efficient environments estimate estimated european examples experimental exploration exposition extend favaro feature features feder figure first fitzgibbon flat foveated freiburg from full generalised generated geometry gibbens given good ground hand harris have http ieee image implementation integrated interest international interplanetary january journal journey june kita kleeman kuniyoshi large learning lenses leonard library linux localisation localization location longitudinally machine made many mapping measured ments method mobile most motion multaneous murray navigation obtained onboard open operating outdoors over overhead oxford pages paper part particularly pattern point position positional positions presented press probabilistic proc proceedings processing ramp ramps real realtime recognition recovery references relevant requiring research returned robot robotics robots rougeaux scale scene seen sequences sequential shaped shown side sifrom simultaneous slam small snowbird soatto society software source springer started starting state stereo structure suehiro surfaces symposium system systems theoretical thesis this through thrun time times together tomasi toward track traversing truth type ultrasonic uncertainty undulating university using utah verlag vertically view vision visual when which while whyte wide with within work yuille zisserman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990478	28	Minimally Supervised Acquisition of 3D Recognition Models from Cluttered Images Andrea Selinger and Randal C. Nelson Department of Computer Science University of Rochester Rochester, NY 14627 fg@cs.rochester.edu selinger,nelson	academy active amini ando appearances architecture automatic barbara basri classification clustering computer conference constraints contours cvpr dimensional dynamic energy erdos evolution fujita gdalyahu graphs hard hierarchical hungarian iccv ieee institute international isbn jacobs june learning mathematical minimizing modular network networks neural objects pages pattern presence proceedings programming publications random recognition references roth santa science silhouettes society suzuki tehrani three unsupervised using vision visual weinshall weymouth
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990643	133	AComputerVisionSystemforOn-ScreenItemSelectionbyFinger Pointing	acombinedcornerandedge aline alvey anandan andapplications andapplicationto andd andinterpretinggestureswithinthecontextofanintelaaaifallsymposiumonembodiedlanguageandaction andk andpelletier andt andthomashuang aparallaxbasedapproach avirtual biologist board braortetal bypointingwithuncalibratedstereovision cipollaandn colmenarez comput computer conference cvpr davitorialobo dblack detectionand detector directrecoveryof dngerdetectionusingasinglecamera dreconstructionfromperspectiveviews finger framestoreferenceplanes freiburg frey fromreference fukumoto germany gestural graph grefit handpostures hanna harrisandm hollinghurst huang human icafgr icip ieee inimageand inproc interfacetoavisualcomputingenvironmentformolecular irani isbn june kumar lnai lyons mase multi navab nolkerandh pages pattern pavlovic pointer pointinginterfacebyimageprocessing proc proceedings recognition references relativeanestructure ritter robotinterface scancomputer seattle shah shapefrommultipleviews sharma shashuaandn society stephens suenaga theccv theory thicpr thusonecameraisalmostcertainlysucient trackingoffacesandfacialfeatures viewparallaxgeometry vision visionalgorithmforidentifyinghumanbodyfeatures visioncomputing visualrecognitionof weinshall
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990529	77	Geometric Blur for Template Matching Alexander C. Berg Jitendra Malik Computer Science Division University of California, Berkeley Abstract 1.1 Related Work	adaptive algorithm america analogues anandan applied based because belongie bergen best better between blur borgefors bridge cars cell cells chamfer computer conference consistent constraint context corners correct correlation correspondences cortex cortical cross cvpr detail detected detection differentiates each eccv edge errors estimation experiment faces feature field figure filtering fine flow found from geometric global hanna have heeger hierarchical hingorani huang iccv ieee ijcai image images incorrect independently interest isbn journal june kanade lack large local localization makes malik many matched matching model models motion multi mumford natural neurosci normalization normalized note object obvious okutomi operator optical over pami parametric pattern points poor proceedings properties provide provides puzicha recognition references regularization relations relationship response responses resulting results rough scale schneiderman september shape sizes small smoothness society space statistical statistics stereo striate support target techniques that them theory there this tradeoff used using vision where window windows with witkin wrong
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990660	149	Structure and Motion Estimation with Expectation Maximization and Extended Kalman Smoother for Continuous Image Sequences	absolute academic acknowledgments adjustment advances affine aiaa alamitos algorithm algorithms analysis anders andrew applicatio applications applied astrom augmented auto autocalibration automatic azarbayejani based batch beal beardsley been berger berthilsson between boston brain bundle calibation calibrated calibration california cambridge camera cipolla cohen cohn collineations computation computer computing conf conference conjunction constant constraints control cornelis cvpr data david davison dempster dept dimensional dublin dynamic dynamical eccv editor editors elsevier embedding engineering environments equation estimates estimation euclidean european evaluation expectation ezio factorization faugeras fitzgibbon focal foundation framework freitas from gelb geometric ghahramani gool graphics gunnar handling hartley heyden hierachy hinton hong icpr ieee image images incomplete india information intelligence internal international intrinsic ireland isbn iterative journal kanatani kearns kita koch korea kosef kurt laird large learning lecture length lengths lens lepetit likelihood linear localisation maarten machine magazine malis mapbuilding marc markerless mass matsunaga maximization maximum maybeck mclauchlan mendonc method metric model models modern monocular moon motion moving multi multiple mumbai murray nando networks neural niste nonlinear notes occlusion optimal optimization ouel oxford pages parameter parameters patter pattern paul paulo pentland philip planar pollefeys practice press proc proceedings processing proesmans projective quadric rauch reality reccursive recognition reconstruction recovery recursive references registeration reliability rendering report research rikard roberto robotics royal rubin scale scene scenes science second selection self semi sequences sequential series signal simon simple society solla sparr spite springer stabilization statistical stochastic striebel structure structures struture supported symposium synthesis systems technical technique tensors theory this three through todd toronto tracking transactions trifocal triggs tung uncalibrated unifying university unknown updating using variable varying vergauwen verlag video view virtual vision vlsi volume weakly weng with workshop yongduek zisserman zoom
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990514	62	Improving the Scope of Deformable Model Shape and Motion Estimation	active advantage aggarwal algebraic aloimonos angular applications approach artificial aspect azarbayejani bajcsy based bike body brown cambridge case citls closed combining computation computer conference constraints contour correspondences crowley cviu cvpr darpa decarlo december deformable deformations density determining dickinson distributed dynamic edge elsevier estimation face faugeras feature feed figure fixture flat flow focal form forward frame from functions geometric geometry global grapple green grey hard hartley horn horowitz huang human iccv ieee ijcv illumination image images implicit incorporating information integrating integration intelligence invariance isbn january july kanantani kang kass kluwer kriegman least leclerc length line linear lines local matches matching measurements metaxas mitiche model modeling models motion mpls multi multiple nato netravali optical optimization orientation pami parametric pattern pentland perception physically physics points practice press probability proc proceedings pugget qualitative quantitative range recognition reconstructed reconstruction recover recovering recovery recursive references relative review rigid rosenfeld samaras scene sclaroff segments seitz sequence series shaded shape shashua skordas snakes society solina solutions spestakis squares statistical stelmaszyk straight streams structure structures superquadrics surfaces szeliski taking taylor tensor terzopoulos theory three tracking trifocal understanding university using vieville view views vision visual wireframe with witkin workshop yellow zisserman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990545	91	1	about acceptable accurate addition adjustment adjustments after algorithm algorithms allggoorriitthhmmss allgoriithm also analysis anndd annotations approach architecture arrowhead average baasseedd base based bbaasseedd berlin both bound challenges chen class compared compariiso complete computer conference coommppaarriissoonn correct corresponding cvpr denotes described detecting detection developed dimension dimensions document doermann dori drawiin drawiing drawing drawings drraawwiinng drraawwiinngg during engineering entity evaluation factor failure false figure frame frames from frraammee further generally geometric geometry germany global graphics grid have higher horizontal however ieee image implemented improved individual integration international introduction isbn kasturi langrana large letters line lines lists loloww machine match mechanical method methods miss missed module most noorrmmaall normal normall number pami parameter parameters pattern percentage performance pixel post presented previously proceedings process processing proposed provide quality qualliit quuaalliittyy raster rate rates ratio recognition recognized recognizing reconstruction rectification references relative representation results rules runtime samal same scale scanned second segmentation seth sets shown size society song springer state strengthening success summary symbol symboll symbols syntactic system systems syymmbbooll taabbllee table tabll tang than that this ting tombre total transactions transformation understanding used vector vectorization vectorized verification verlag vertical vision were with
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990455	6	Efficient Evaluation of Classification and Recognition Systems	academic against algorithms american analaysis analysis annals approach assessment association binford biometrika blackburn bone bootstrap boult bowyer burman cabrera cambridge chapter char characterization chateristics ciang classification cochran complex computer conference cons construction crafters cvgip cvpr data design dodcounterdrug duda edition editors efficient empirical estimation estimators evaluation experiments face facial facialrecognition feret frankel from government gurney hart http hyeonjoon ieee ignorance image imaging inference intelligence introduction investication isbn jain jewett john jonathon june kish klette kluwer krewski lehigh lehtonen litho machine martin matei mccarthy medical meer methodology methods micheals moon motion multifactorial myopia naivete october office optimum orthogonal pahkinen pami patrick pattern performance phillips plackett practical practice printing proceedings pros publishers rauss recognition references replicate replication replications report resampling research resoluation resolution rigid rizvi robert ross royal rust samples sampling second society sons springer stat statified statistical statistics stiehl stner stork super survey surveys syed system systems technical techniques terrance test third through transactions tyler understanding university using variance vendor verlag viergever vincken vision warping wiley wolter work
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990465	16	Simultaneous linear estimation of multiple view geometry and lens distortion Andrew W Fitzgibbon Department of Engineering Science The University of Oxford 19 Parks Road, Oxford, United Kingdom, OX1 3PJ	active adjustment again algorithm alignment allows allsolutions although american analysis andr angle appendix application applied automatic baltimore because been begin between block bmvc boujou brady brown bundle calibration cambridge cameras case cause center change characterization church columns computations computer conference constraints convert correction correspond correspondences cvpr czech decentering define degeneracies devernay diego digital distortion does drop edition editors eigen eigensystem eigenvalue eigenvalues eight engineering environments epipolar equations falls faugeras fewer fitzgibbon follows form found from further future general generalized geometric geometry giving golub hartley have hichcorrespondtonon hopkins http iccv icpr ieee image images imaginary include indefenceofthe infinite inpractice inproc intrinsic isbn ization jaenicke johns kanatani kang kumar least lens lenses lncs loan make manual mathods matrices matrix mclauchlan meerbergen minimum modern more mosaicing most motion multi multiple nayar nonmetric nullvectors number obtain obviously ofthese original pages pair pami parameters pattern patterns photogrammetric photogrammetry photometric point pointalgorithm points practice prague premultiplication premultiplying press problem proc proceedings proof quadratic radial rank recognition recovering reduced references removal report republic review satisfy sawhney scenes second self semiautomatic sequential shafer shows siam simultaneously single singular slama society soluniques solution solutions solve solved solvers solving spie springer square squares standard statisticalbiasofconicfittingandrenormal stein structure structured sturm swaminathan synthesis system systems szeliski tech technical than theory ther these thesis they this those three thuswemaysolvethe tion tisseur triangulation triggs true ular university unsymmetric used using values variable verlag view views vision visionalgorithms volume what which wide will willson with work written yield zero zhang zisserman
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990463	14	Rectifying Transformations That Minimize Resampling Effects	aloimonos amount applications approximate area arnold artificial ayache binocular biological camera change class classical computation computer computing conf conference configurations considered created cvpr cylindrical dennis deriche derived dimensional discussed distortion editor efficient encountered epipolar estimation faugeras first from general geometric geometries geometry given gool ground guided hansen hartley hebert homographies however ieee image images intelligence international isbn journal knowledge koch lastly line local loop lost luong matching mathematical mechanics method methods metric meunier mines minimize minimizes model most motion navigation need only optimal original pages pair pairs papadimitriou parameterization parameters pattern perspective pixels pollefeys practice preferable preserving press proc proceedings processing projective proposed recognition recovering recovery rectification rectify rectifying references robert robotics robust sampling scale showed simple society solution some springer stereo stereovision structure such sufficient systems tasks technique that then theory this those three through trans transform transforms trinocular uncalibrated unknown unmanned using value values vehicles verlag viewpoint vision visual visually when while zeller zhang
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990629	122	A Weighted Non-negative Matrix Factorization for Local Representations David Guillamet, Marco Bressan and Jordi Vitria` Centre de Visio´ per Computador, Dept. Informa`tica Universitat Autonoma` de Barcelona, 08193 Bellaterra, Barcelona, Spain. davidg, marco, jordi @cvc.uab.es	ability according account algorithm algorithms also analysis appearance approach approaches articulated based basis because behaviours better black both caip california capacity classical classification closer clusters cognitive compared component components computer conf conference considered considering contain cuts cvpr data database demonstrate desired developed development different dimensionality discriminant done each eigenfaces eigentracking eliminated error experiment fact factorization finally find finds found from further guillamet have high however iccv iciap ieee ijcv image implies improving increased information intern international into isbn jepson journal learning level lncs local malik matching matrix method murase nature nayar negative neuroscience nips nonnegative normalized number numerical object objects obtain obtained only original other pages pami partbased parts pattern pentland performance previously primitives principal proc proceedings processing projected recognition reconstruction reduce reduction redundancy redundant references representation representations representing respect robust seen segmentation sense seung shams shape shows society some sort southern specific such sults superior takes techniques terms tests that thesis they this torre tracking traditional training trans turk university unsupervised using vector view vision visual vitria weight weighted when with
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990556	101	Robust Change Detection by Fusing Intensity and Texture Differences	accurate achieved acknowledgements activity adaptive addison also analyzed authors automatic azarbaygaui background based been better bichsel body both brumitt buxton cgit chang change changes color combination combining commun complex complexity computational computer conference connected currently cvgip cvpr cvprip darrell derived detection developed difference differences duric edge energy environments error evaluations examples experiment extraction facilities feature ferences figure filters from fusion future gabor generated global glossary gradient grimson have hence hill human iccv icpr ieee illumination image images implementation impressive improvement independent information integrate integrating integration intensity into investigated isbn jabri jain kasturi krumm learning leung like likelihood location machine maintenance matrix mcgraw measure measures methods meyers minimized moments more motion moving murray noise objects occurrence optimization pages pami parameters pattern patterns pentland people pfinder phong pictures practice presented press principles proc proceedings processing proposed providing quantitative real reality recent recognition reed references relation respect results review robust robustness rosenfeld rosin scene schunck segmentation segmenting selection sequences should showed simply skifstad society standard static statistical stauffer studies such surveillance systems technique techniques terminology test texture thank that this thresholding time toyama tracking trans under understanding upon using varying vectors video vince virtual vision visual wallflower wavelet wechsler weighting were wesley with wokingham world would wren york
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990450	1	Small Sample Learning during Multimedia Retrieval using BiasMap Xiang Sean Zhou, Thomas S. Huang Beckman Institute, University of Illinois at Urbana Champaign	academic access advances algorithm amer analysis anouar approach asso austin based baudat boosting boston carolina chang chen class classification color comp computation computer conf conference content cvpr databases discriminant discrimination duda edge examples exploring faloutsos feature features feedback fisher friedman fukunaga generalized greece hart hawaii head hilton huang hunag icip ieee image images information introduction isbn ishikawa island john jose june kernel kinh large learning letters libraries mathematical mika mindreader muller multiple nature neural optimizing orengo pattern paul press proc proceedings processing programming query ratsch recognition references regularized relevance retrieval scene similarity smith society sons south spie springer stat statistical stor stricker structural subramanya systems texture theory through tieu transform using vapnik variants video viola vision vldb wiley workshop york zhou
http://doi.ieeecomputersociety.org/10.1109/CVPR.2001.990483	32	Efficient Grouping under Perspective Skew Andreas Turina1 Tinne Tuytelaars2 Luc Van Gool1	acknowledge acknowledgments advances aerial affinely algebraic also analysis analyzer applications automated automatic based baseline been between bolles boult british cartography cascaded cham cipolla collins commun computational computer computing concept conf conference consensus contours correspondences curve cvpr dealt detecting detection dhome different elements engineering european finally fischler fitting fixed flat frieze from geometric geometry glachet gool graphics gratefully gross grouping groupings groups have hough ieee image interpretation interrelations intl invariant isbn kluwer kneebone lapreste leung lines local localizing locating lowe machine malik matching mathematical model modeling moons object organization oxford paradigm pattern patterns perception perceptual philosophical physical planar points present press proc proceedings processing proesmans project projective random recognition references regions repeated reveals royal saliency sample scene schaffalitzky sciences semple separately single society some stereo structures subgroups support syman symmetric symmetry that traditionally transactions transform turina tuytelaars types understanding university using vanishing vibes vision visual wallpaper wide with zisserman zuri